Search Results

Search found 406 results on 17 pages for 'forgotten'.

Page 16/17 | < Previous Page | 12 13 14 15 16 17  | Next Page >

  • Database Schema Versioning Strategies

    - by Jack Ryan
    I work on a project that uses a reasonably large database, the live version weighing in at somewhere around 60-80GB. The live database is the only real definitive source of our schema, and because of its size duplicating this database is too slow to be done often. This means we have ended up developing our database schema in a pretty ad hoc way, using sql compare to migrate changes from dev dbs to the live system, and only wiping our dev dbs every month or two. I am hoping to get some pointers on how to improve our database development work flow so that we have a little more control. Some things to think about: Currently nobody is really in charge of the database schema, all developers can change it if they need to, though generally these decisions are talked about before they are done. There are stored procedures, functions, and views in the database. These should probably be dumped to files so they can be reloaded on every build. Schema changes should probably be checked in as scripts. We have started to do this recently. However all our scripts must then be numbered (because there may be dependencies between them), and must be re runnable (because our build script currently runs them all in order). This makes them hard to read because they are full of conditionals that check whether tables or columns already exist. This is a step that is often forgotten by developers. Getting a new database should be quick and easy. This is currently a big problem, it takes several hours to get a copy of last nights backup and restore it onto a dev machine. Some mechanism needs to be in place to allow developers to update static data. We have tables that contain data that is never updated through the application, but does potentially need to be changed when we do a new release (often this drives dropdowns). The whole thing needs to be runnable as part of a build script. Are there any tools that can be used to help to do this? Eventually I would like to be at a point where a new DB can be built from scratch without copying any data from the live system. I don't mind writing some scripts to glue all the steps together but each part should be easily editable so that we continue to use it rather than make changes directly on DBs.

    Read the article

  • MySQL my.cnf file not being read, Ubuntu 10.04 64bit

    - by reallyordinary
    I've been researching this for a few hours with no luck. Basically it looks like my server's my.cnf file isn't being read at all. I've searched my server, and there's only one my.cnf file on it, located at /etc/mysql/my.cnf. Its ownership is root:root. I'm running Ubuntu 10.04 64bit on a Linode.com server. I have the latest versions of MySQL and PHP installed. I've edited the my.cnf file, commented out "skip-innodb", and have set innodb to be the default storage engine using default-storage-engine = innodb And then restarted mysql. But when I do show engines, MyISAM is still coming up as the default engine. Also - none of the innodb settings I've added to the my.cnf file are being read. For example, I have this in my.cnf: innodb_buffer_pool_size=4G But in phpmyadmin, InnoDB is showing as having a buffer pool size of 8,192 KiB. Similarly, I have this in the my.cnf: innodb_data-file_path = ibdata1:500M:autoextend But in phpmyadmin, it's reading as ibdata1:10M:autoextend. It doesn't look like MyISAM info is being read from the my.cnf file either. The my.cnf file has skip-external-locking queried out, but it's showing as "on" in phpmyadmin. So - yeah, it looks like nothing in the my.cnf file is being read at all. But the server still works. I'm running a Drupal site on it and it seems to operate fine. So mysql seems to be drawing default settings from... some mysterious secret location. Any idea how I can make mysql see and use this my.cnf file? Actually, wait - it looks like it may be being read, not sure. I checked the error.log and found this: 101128 4:28:52 [ERROR] Cannot find or open table databasename/cache_apachesolr from the internal data dictionary of InnoDB though the .frm file for the table exists. Maybe you have deleted and recreated InnoDB data files but have forgotten to delete the corresponding .frm files of InnoDB tables, or you have moved .frm files to another database? or, the table contains indexes that this version of the engine doesn't support. See http://dev.mysql.com/doc/refman/5.1/en/innodb-troubleshooting.html how you can resolve the problem. InnoDB: Error: auto-extending data file ./ibdata1 is of a different size InnoDB: 640 pages (rounded down to MB) than specified in the .cnf file: InnoDB: initial 32000 pages, max 0 (relevant if non-zero) pages! InnoDB: Could not open or create data files. InnoDB: If you tried to add new data files, and it failed here, InnoDB: you should now edit innodb_data_file_path in my.cnf back InnoDB: to what it was, and remove the new ibdata files InnoDB created InnoDB: in this failed attempt. InnoDB only wrote those files full of InnoDB: zeros, but did not yet use them in any way. But be careful: do not InnoDB: remove old data files which contain your precious data! 101128 4:28:52 [ERROR] Plugin 'InnoDB' init function returned error. 101128 4:28:52 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed. 101128 4:28:52 [ERROR] /usr/sbin/mysqld: unknown variable 'innodb_lock_wait_timout=50' 101128 4:28:52 [ERROR] Aborting 101128 4:28:52 [Note] /usr/sbin/mysqld: Shutdown complete

    Read the article

  • Replicate between mysql 5.0.xx community and enterprise edition over ssh

    - by Arlukin
    I'm trying to setup a mysql replication over an SSH tunnel. The odd thing about this setup is that I have one master with mysql 5.0.60sp1-enterprise-gpl-log and one slave with mysql 5.0.67-community-log. Could it be so that it's not possible to replicate between community and enterprise edition? As you can see in my log below, it's possible to login on the remote server with the mysql client. But the replication get "Can't connect to MySQL server on '127.0.0.1' (13)" Is it any log file I have forgotten to look in, to get more info? [root@mysql1-av ~]# mysql -uroot -p Enter password: Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 73 Server version: 5.0.67-community-log MySQL Community Edition (GPL) Type 'help;' or '\h' for help. Type '\c' to clear the buffer. The version of the slave mysql [root@mysql1-av ~]# autossh -f -M 20001 -L 3307:10.200.200.200:3306 [email protected] -N [root@mysql1-av ~]# mysql -h127.0.0.1 --port 3307 -uroot -p Enter password: Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 5189 Server version: 5.0.60sp1-enterprise-gpl-log MySQL Enterprise Server (GPL) Type 'help;' or '\h' for help. Type '\c' to clear the buffer. mysql> Aborted Login to the master mysql with the mysql client over the ssh tunnel. [root@mysql1-av ~]# mysql -uroot -p Enter password: Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 75 Server version: 5.0.67-community-log MySQL Community Edition (GPL) Type 'help;' or '\h' for help. Type '\c' to clear the buffer. mysql> change master to master_host='127.0.0.1', MASTER_PORT=3307, master_user='xxxx', master_password='xxxx', master_log_file='bin.000001'; Query OK, 0 rows affected (0.00 sec) mysql> start slave; Query OK, 0 rows affected (0.00 sec) mysql> show slave status \G *************************** 1. row *************************** Slave_IO_State: Connecting to master Master_Host: 127.0.0.1 Master_User: replNSG Master_Port: 3307 Connect_Retry: 60 Master_Log_File: bin.000001 Read_Master_Log_Pos: 4 Relay_Log_File: relay.000001 Relay_Log_Pos: 98 Relay_Master_Log_File: bin.000001 Slave_IO_Running: No Slave_SQL_Running: Yes Replicate_Do_DB: Replicate_Ignore_DB: Replicate_Do_Table: Replicate_Ignore_Table: Replicate_Wild_Do_Table: Replicate_Wild_Ignore_Table: Last_Errno: 0 Last_Error: Skip_Counter: 0 Exec_Master_Log_Pos: 4 Relay_Log_Space: 98 Until_Condition: None Until_Log_File: Until_Log_Pos: 0 Master_SSL_Allowed: No Master_SSL_CA_File: Master_SSL_CA_Path: Master_SSL_Cert: Master_SSL_Cipher: Master_SSL_Key: Seconds_Behind_Master: NULL 1 row in set (0.00 sec) Start the replication, but it breaks on IO. [root@mysql1-av ~]# tail /var/log/mysqld.log 120921 22:17:59 [Note] Slave I/O thread killed while connecting to master 120921 22:17:59 [Note] Slave I/O thread exiting, read up to log 'bin.000001', position 4 120921 22:17:59 [Note] Error reading relay log event: slave SQL thread was killed 120921 22:29:36 [Note] Slave SQL thread initialized, starting replication in log 'bin.000001' at position 4, relay log '/var/lib/mysql/relay.000001' position: 4 120921 22:29:36 [ERROR] Slave I/O thread: error connecting to master '[email protected]:3307': Error: 'Can't connect to MySQL server on '127.0.0.1' (13)' errno: 2003 retry-time: 60 retries: 86400 Because it can't connect to the master server.

    Read the article

  • How can I make subversion reset the stored passwords/users and remember my authentication credential

    - by NicDumZ
    Hello folks! Background: I used to have everything working just fine on my fresh install: $ svn co https://domain:443/ test1 Error validating server certificate for 'https://domain:443': - The certificate is not issued by a trusted authority. Use the fingerprint to validate the certificate manually! Certificate information: - Hostname: **REMOVED** - Valid: **REMOVED** - Issuer: **REMOVED** - Fingerprint: **checked with issuer and REMOVED** (R)eject, accept (t)emporarily or accept (p)ermanently? p Authentication realm: <https://domain:443> Subversion repository Password for 'nicdumz-machine-hostname': Authentication realm: <https://domain:443> Subversion repository Username: nicdumz Password for 'nicdumz': # proceeds to checkout correctly $ svn co https://domain:443/ test2 # checkouts nicely, without asking for my password. At some point I needed to commit stuff using a different account. So I did that $ svn ci --username other.user Authentication realm: <https://domain:443> Subversion repository Password for 'other.user': # works fine But since then, everytime I want to commit as 'nicdumz' (default user, all repos have been checked-out with that user), it prompts me for my password: $ svn ci Authentication realm: <https://domain:443> Subversion repository Password for 'nicdumz': Hey come on, why :) The same happens if I want a fresh checkout, since read-access is also protected. So I tried fixing the issue by myself. I read around that ~/.subversion/auth was storing credentials, so I removed it from the way: $ cd ~/.subversion $ mv auth oldauth $ mkdir auth It seemed to work at first, because svn had forgotten about certificate validation: $ svn co https://domain:443/ test3 Error validating server certificate for 'https://domain:443': - The certificate is not issued by a trusted authority. Use the fingerprint to validate the certificate manually! Certificate information: - Hostname: **REMOVED** - Valid: **REMOVED** - Issuer: **REMOVED** - Fingerprint: **checked with issuer and REMOVED** (R)eject, accept (t)emporarily or accept (p)ermanently? p Authentication realm: <https://domain:443> Subversion repository Password for 'nicdumz-machine-hostname': Authentication realm: <https://domain:443> Subversion repository Username: nicdumz Password for 'nicdumz': # proceeds to checkout correctly $ svn up Authentication realm: <https://domain:443> Subversion repository Password for 'nicdumz': What? how is this happening? If you have suggestions to investigate more about the behaviour, I am very interested. If I'm correct, there is no way to do a verbose svn up or anything of the like, so I'm not sure should I go for investigation. Oh, and for what it's worth: $ svn --version svn, version 1.6.6 (r40053) compiled Oct 26 2009, 06:19:08 Copyright (C) 2000-2009 CollabNet. Subversion is open source software, see http://subversion.tigris.org/ This product includes software developed by CollabNet (http://www.Collab.Net/). The following repository access (RA) modules are available: * ra_neon : Module for accessing a repository via WebDAV protocol using Neon. - handles 'http' scheme - handles 'https' scheme * ra_svn : Module for accessing a repository using the svn network protocol. - with Cyrus SASL authentication - handles 'svn' scheme * ra_local : Module for accessing a repository on local disk. - handles 'file' scheme * ra_serf : Module for accessing a repository via WebDAV protocol using serf. - handles 'http' scheme - handles 'https' scheme

    Read the article

  • Replacing jQuery.live() with jQuery.on()

    - by Rick Strahl
    jQuery 1.9 and 1.10 have introduced a host of changes, but for the most part these changes are mostly transparent to existing application usage of jQuery. After spending some time last week with a few of my projects and going through them with a specific eye for jQuery failures I found that for the most part there wasn't a big issue. The vast majority of code continues to run just fine with either 1.9 or 1.10 (which are supposed to be in sync but with 1.10 removing support for legacy Internet Explorer pre-9.0 versions). However, one particular change in the new versions has caused me quite a bit of update trouble, is the removal of the jQuery.live() function. This is my own fault I suppose - .live() has been deprecated for a while, but with 1.9 and later it was finally removed altogether from jQuery. In the past I had quite a bit of jQuery code that used .live() and it's one of the things that's holding back my upgrade process, although I'm slowly cleaning up my code and switching to the .on() function as the replacement. jQuery.live() jQuery.live() was introduced a long time ago to simplify handling events on matched elements that exist currently on the document and those that are are added in the future and also match the selector. jQuery uses event bubbling, special event binding, plus some magic using meta data attached to a parent level element to check and see if the original target event element matches the selected selected elements (for more info see Elijah Manor's comment below). An Example Assume a list of items like the following in HTML for example and further assume that the items in this list can be appended to at a later point. In this app there's a smallish initial list that loads to start, and as the user scrolls towards the end of the initial small list more items are loaded dynamically and added to the list.<div id="PostItemContainer" class="scrollbox"> <div class="postitem" data-id="4z6qhomm"> <div class="post-icon"></div> <div class="postitemheader"><a href="show/4z6qhomm" target="Content">1999 Buick Century For Sale!</a></div> <div class="postitemprice rightalign">$ 3,500 O.B.O.</div> <div class="smalltext leftalign">Jun. 07 @ 1:06am</div> <div class="post-byline">- Vehicles - Automobiles</div> </div> <div class="postitem" data-id="2jtvuu17"> <div class="postitemheader"><a href="show/2jtvuu17" target="Content">Toyota VAN 1987</a></div> <div class="postitemprice rightalign">$950</div> <div class="smalltext leftalign">Jun. 07 @ 12:29am</div> <div class="post-byline">- Vehicles - Automobiles</div> </div> … </div> With the jQuery.live() function you could easily select elements and hook up a click handler like this:$(".postitem").live("click", function() {...}); Simple and perfectly readable. The behavior of the .live handler generally was the same as the corresponding simple event handlers like .click(), except that you have to explicitly name the event instead of using one of the methods. Re-writing with jQuery.on() With .live() removed in 1.9 and later we have to re-write .live() code above with an alternative. The jQuery documentation points you at the .on() or .delegate() functions to update your code. jQuery.on() is a more generic event handler function, and it's what jQuery uses internally to map the high level event functions like .click(),.change() etc. that jQuery exposes. Using jQuery.on() however is not a one to one replacement of the .live() function. While .on() can handle events directly and use the same syntax as .live() did, you'll find if you simply switch out .live() with .on() that events on not-yet existing elements will not fire. IOW, the key feature of .live() is not working. You can use .on() to get the desired effect however, but you have to change the syntax to explicitly handle the event you're interested in on the container and then provide a filter selector to specify which elements you are actually interested in for handling the event for. Sounds more complicated than it is and it's easier to see with an example. For the list above hooking .postitem clicks, using jQuery.on() looks like this:$("#PostItemContainer").on("click", ".postitem", function() {...}); You specify a container that can handle the .click event and then provide a filter selector to find the child elements that trigger the  the actual event. So here #PostItemContainer contains many .postitems, whose click events I want to handle. Any container will do including document, but I tend to use the container closest to the elements I actually want to handle the events on to minimize the event bubbling that occurs to capture the event. With this code I get the same behavior as with .live() and now as new .postitem elements are added the click events are always available. Sweet. Here's the full event signature for the .on() function: .on( events [, selector ] [, data ], handler(eventObject) ) Note that the selector is optional - if you omit it you essentially create a simple event handler that handles the event directly on the selected object. The filter/child selector required if you want life-like - uh, .live() like behavior to happen. While it's a bit more verbose than what .live() did, .on() provides the same functionality by being more explicit on what your parent container for trapping events is. .on() is good Practice even for ordinary static Element Lists As a side note, it's a good practice to use jQuery.on() or jQuery.delegate() for events in most cases anyway, using this 'container event trapping' syntax. That's because rather than requiring lots of event handlers on each of the child elements (.postitem in the sample above), there's just one event handler on the container, and only when clicked does jQuery drill down to find the matching filter element and tries to match it to the originating element. In the early days of jQuery I used manually build handlers that did this and manually drilled from the event object into the originalTarget to determine if it's a matching element. With later versions of jQuery the various event functions in jQuery essentially provide this functionality out of the box with functions like .on() and .delegate(). All of this is nothing new, but I thought I'd write this up because I have on a few occasions forgotten what exactly was needed to replace the many .live() function calls that litter my code - especially older code. This will be a nice reminder next time I have a memory blank on this topic. And maybe along the way I've helped one or two of you as well to clean up your .live() code…© Rick Strahl, West Wind Technologies, 2005-2013Posted in jQuery   Tweet !function(d,s,id){var js,fjs=d.getElementsByTagName(s)[0];if(!d.getElementById(id)){js=d.createElement(s);js.id=id;js.src="//platform.twitter.com/widgets.js";fjs.parentNode.insertBefore(js,fjs);}}(document,"script","twitter-wjs"); (function() { var po = document.createElement('script'); po.type = 'text/javascript'; po.async = true; po.src = 'https://apis.google.com/js/plusone.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(po, s); })();

    Read the article

  • Fragmented Log files could be slowing down your database

    - by Fatherjack
    Something that is sometimes forgotten by a lot of DBAs is the fact that database log files get fragmented in the same way that you get fragmentation in a data file. The cause is very different but the effect is the same – too much effort reading and writing data. Data files get fragmented as data is changed through normal system activity, INSERTs, UPDATEs and DELETEs cause fragmentation and most experienced DBAs are monitoring their indexes for fragmentation and dealing with it accordingly. However, you don’t hear about so many working on their log files. How can a log file get fragmented? I’m glad you asked. When you create a database there are at least two files created on the disk storage; an mdf for the data and an ldf for the log file (you can also have ndf files for extra data storage but that’s off topic for now). It is wholly possible to have more than one log file but in most cases there is little point in creating more than one as the log file is written to in a ‘wrap-around’ method (more on that later). When a log file is created at the time that a database is created the file is actually sub divided into a number of virtual log files (VLFs). The number and size of these VLFs depends on the size chosen for the log file. VLFs are also created in the space added to a log file when a log file growth event takes place. Do you have your log files set to auto grow? Then you have potentially been introducing many VLFs into your log file. Let’s get to see how many VLFs we have in a brand new database. USE master GO CREATE DATABASE VLF_Test ON ( NAME = VLF_Test, FILENAME = 'C:\Program Files\Microsoft SQL Server\MSSQL10.ROCK_2008\MSSQL\DATA\VLF_Test.mdf', SIZE = 100, MAXSIZE = 500, FILEGROWTH = 50 ) LOG ON ( NAME = VLF_Test_Log, FILENAME = 'C:\Program Files\Microsoft SQL Server\MSSQL10.ROCK_2008\MSSQL\DATA\VLF_Test_log.ldf', SIZE = 5MB, MAXSIZE = 250MB, FILEGROWTH = 5MB ); go USE VLF_Test go DBCC LOGINFO; The results of this are firstly a new database is created with specified files sizes and the the DBCC LOGINFO results are returned to the script editor. The DBCC LOGINFO results have plenty of interesting information in them but lets first note there are 4 rows of information, this relates to the fact that 4 VLFs have been created in the log file. The values in the FileSize column are the sizes of each VLF in bytes, you will see that the last one to be created is slightly larger than the others. So, a 5MB log file has 4 VLFs of roughly 1.25 MB. Lets alter the CREATE DATABASE script to create a log file that’s a bit bigger and see what happens. Alter the code above so that the log file details are replaced by LOG ON ( NAME = VLF_Test_Log, FILENAME = 'C:\Program Files\Microsoft SQL Server\MSSQL10.ROCK_2008\MSSQL\DATA\VLF_Test_log.ldf', SIZE = 1GB, MAXSIZE = 25GB, FILEGROWTH = 1GB ); With a bigger log file specified we get more VLFs What if we make it bigger again? LOG ON ( NAME = VLF_Test_Log, FILENAME = 'C:\Program Files\Microsoft SQL Server\MSSQL10.ROCK_2008\MSSQL\DATA\VLF_Test_log.ldf', SIZE = 5GB, MAXSIZE = 250GB, FILEGROWTH = 5GB ); This time we see more VLFs are created within our log file. We now have our 5GB log file comprised of 16 files of 320MB each. In fact these sizes fall into all the ranges that control the VLF creation criteria – what a coincidence! The rules that are followed when a log file is created or has it’s size increased are pretty basic. If the file growth is lower than 64MB then 4 VLFs are created If the growth is between 64MB and 1GB then 8 VLFs are created If the growth is greater than 1GB then 16 VLFs are created. Now the potential for chaos comes if the default values and settings for log file growth are used. By default a database log file gets a 1MB log file with unlimited growth in steps of 10%. The database we just created is 6 MB, let’s add some data and see what happens. USE vlf_test go -- we need somewhere to put the data so, a table is in order IF OBJECT_ID('A_Table') IS NOT NULL DROP TABLE A_Table go CREATE TABLE A_Table ( Col_A int IDENTITY, Col_B CHAR(8000) ) GO -- Let's check the state of the log file -- 4 VLFs found EXECUTE ('DBCC LOGINFO'); go -- We can go ahead and insert some data and then check the state of the log file again INSERT A_Table (col_b) SELECT TOP 500 REPLICATE('a',2000) FROM sys.columns AS sc, sys.columns AS sc2 GO -- insert 500 rows and we get 22 VLFs EXECUTE ('DBCC LOGINFO'); go -- Let's insert more rows INSERT A_Table (col_b) SELECT TOP 2000 REPLICATE('a',2000) FROM sys.columns AS sc, sys.columns AS sc2 GO 10 -- insert 2000 rows, in 10 batches and we suddenly have 107 VLFs EXECUTE ('DBCC LOGINFO'); Well, that escalated quickly! Our log file is split, internally, into 107 fragments after a few thousand inserts. The same happens with any logged transactions, I just chose to illustrate this with INSERTs. Having too many VLFs can cause performance degradation at times of database start up, log backup and log restore operations so it’s well worth keeping a check on this property. How do we prevent excessive VLF creation? Creating the database with larger files and also with larger growth steps and actively choosing to grow your databases rather than leaving it to the Auto Grow event can make sure that the growths are made with a size that is optimal. How do we resolve a situation of a database with too many VLFs? This process needs to be done when the database is under little or no stress so that you don’t affect system users. The steps are: BACKUP LOG YourDBName TO YourBackupDestinationOfChoice Shrink the log file to its smallest possible size DBCC SHRINKFILE(FileNameOfTLogHere, TRUNCATEONLY) * Re-size the log file to the size you want it to, taking in to account your expected needs for the coming months or year. ALTER DATABASE YourDBName MODIFY FILE ( NAME = FileNameOfTLogHere, SIZE = TheSizeYouWantItToBeIn_MB) * – If you don’t know the file name of your log file then run sp_helpfile while you are connected to the database that you want to work on and you will get the details you need. The resize step can take quite a while This is already detailed far better than I can explain it by Kimberley Tripp in her blog 8-Steps-to-better-Transaction-Log-throughput.aspx. The result of this will be a log file with a VLF count according to the bullet list above. Knowing when VLFs are being created By complete coincidence while I have been writing this blog (it’s been quite some time from it’s inception to going live) Jonathan Kehayias from SQLSkills.com has written a great article on how to track database file growth using Event Notifications and Service Broker. I strongly recommend taking a look at it as this is going to catch any sneaky auto grows that take place and let you know about them right away. Hassle free monitoring of VLFs If you are lucky or wise enough to be using SQL Monitor or another monitoring tool that let’s you write your own custom metrics then you can keep an eye on this very easily. There is a custom metric for VLFs (written by Stuart Ainsworth) already on the site and there are some others there are very useful so take a moment or two to look around while you are there. Resources MSDN – http://msdn.microsoft.com/en-us/library/ms179355(v=sql.105).aspx Kimberly Tripp from SQLSkills.com – http://www.sqlskills.com/BLOGS/KIMBERLY/post/8-Steps-to-better-Transaction-Log-throughput.aspx Thomas LaRock at Simple-Talk.com – http://www.simple-talk.com/sql/database-administration/monitoring-sql-server-virtual-log-file-fragmentation/ Disclosure I am a Friend of Red Gate. This means that I am more than likely to say good things about Red Gate DBA and Developer tools. No matter how awesome I make them sound, take the time to compare them with other products before you contact the Red Gate sales team to make your order.

    Read the article

  • Video on Architecture and Code Quality using Visual Studio 2012&ndash;interview with Marcel de Vries and Terje Sandstrom by Adam Cogan

    - by terje
    Find the video HERE. Adam Cogan did a great Web TV interview with Marcel de Vries and myself on the topics of architecture and code quality.  It was real fun participating in this session.  Although we know each other from the MVP ALM community,  Marcel, Adam and I haven’t worked together before. It was very interesting to see how we agreed on so many terms, and how alike we where thinking.  The basics of ensuring you have a good architecture and how you could document it is one thing.  Also, the same agreement on the importance of having a high quality code base, and how we used the Visual Studio 2012 tools, and some others (NDepend for example)  to measure and ensure that the code quality was where it should be.  As the tools, methods and thinking popped up during the interview it was a lot of “Hey !  I do that too!”.  The tools are not only for “after the fact” work, but we use them during the coding.  That way the tools becomes an integrated part of our coding work, and helps us to find issues we may have overlooked.  The video has a bunch of call outs, pinpointing important things to remember. These are also listed on the corresponding web page. I haven’t seen that touch before, but really liked this way of doing it – it makes it much easier to spot the highlights.  Titus Maclaren and Raj Dhatt from SSW have done a terrific job producing this video.  And thanks to Lei Xu for doing the camera and recording job.  Thanks guys ! Also, if you are at TechEd Amsterdam 2012, go and listen to Adam Cogan in his session on “A modern architecture review: Using the new code review tools” Friday 29th, 10.15-11.30 and Marcel de Vries session on “Intellitrace, what is it and how can I use it to my benefit” Wednesday 27th, 5-6.15 The highlights points out some important practices.  I’ll elaborate on a few of them here: Add instructions on how to compile the solution.  You do this by adding a text file with instructions to the solution, and keep it under source control.  These instructions should contain what is needed on top of a standard install of Visual Studio.  I do a lot of code reviews, and more often that not, I am not even able to compile the program, because they have used some tool or library that needs to be installed.  The same applies to any new developer who enters into the team, so do this to increase your productivity when the team changes, or a team member switches computer. Don’t forget to document what you have to configure on the computer, the IIS being a common one. The more automatic you can do this, the better.  Use NuGet to get down libraries. When the text document gets more than say, half a page, with a bunch of different things to do, convert it into a powershell script instead.  The metrics warning levels.  These are very conservatively set by Microsoft.  You rarely see anything but green, and besides, you should have color scales for each of the metrics.  I have a blog post describing a more appropriate set of levels, based on both research work and industry “best practices”.  The essential limits are: Cyclomatic complexity and coupling:  Higher numbers are worse On method levels: Green :  From 0 to 10 Yellow:  From 10 to 20  (some say 15).   Acceptable, but have a look to see if there is something unneeded here. Red: From 20 to 40:   Action required, get these down. Bleeding Red: Above 40   This is the real red alert.  Immediate action!  (My invention, as people have asked what do I do when I have cyclomatic complexity of 150.  The only answer I could think of was: RUN! ) Maintainability index:  Lower numbers are worse, scale from 0 to 100. On method levels: Green:  60 to 100 Yellow:  40 – 60.    You will always have methods here too, accept the higher ones, take a look at those who are down to the lower limit.  Check up against the other metrics.) Red:  20 – 40:  Action required, fix these. Bleeding red:  Below 20.  Immediate action required. When doing metrics analysis, you should leave the generated code out.  You do this by adding attributes, unfortunately Microsoft has “forgotten” to add these to all their stuff, so you might have to add them to some of the code.  It most cases it can be done so that it is not overwritten by a new round of code generation.  Take a look a my blog post here for details on how to do that. Class level metrics might also be useful, at least for coupling and maintenance.  But it is much more difficult to set any fixed limits on those.  Any metric aggregations on higher level tend to be pretty useless, as the number of methods vary pretty much, and there are little science on what number of methods can be regarded as good or bad.  NDepend have a recommendation, but they say it may vary too.  And in these days of data binding, the number might be pretty high, as properties counts as methods.  However, if you take the worst case situations, classes with more than 20 methods are suspicious, and coupling and cyclomatic complexity go red above 20, so any classes with more than 20x20 = 400 for these measures should be checked over. In the video we mention the SOLID principles, coined by “Uncle Bob” (Richard Martin). One of them, the Dependency Inversion principle we discuss in the video.  It is important to note that this principle is NOT on whether you should use a Dependency Inversion Container or not, it is about how you design the interfaces and interactions between your classes.  The Dependency Inversion Container is just one technique which is based on this principle, but which main purpose is to isolate things you would like to change at runtime, for example if you implement a plug in architecture.  Overuse of a Dependency Inversion Container is however, NOT a good thing.  It should be used for a purpose and not as a general DI solution.  The general DI solution and thinking however is useful far beyond the DIC.   You should always “program to an abstraction”, and not to the concreteness.  We also talk a bit about the GRASP patterns, a term coined by Craig Larman in his book Applying UML and design patterns. GRASP patterns stand for General Responsibility Assignment Software Patterns and describe fundamental principles of object design and responsibility assignment.  What I find great with these patterns is that they is another way to focus on the responsibility of a class.  One of the things I most often found that is broken in software designs, is that the class lack responsibility, and as a result there are a lot of classes mucking around in the internals of the other classes.  We also discuss the term “Code Smells”.  This term was invented by Kent Beck and Martin Fowler when they worked with Fowler’s “Refactoring” book. A code smell is a set of “bad” coding practices, which are the drivers behind a corresponding set of refactorings.  Here is a good list of the smells, and their corresponding refactor patterns. See also this.

    Read the article

  • Scripting Language Sessions at Oracle OpenWorld and MySQL Connect, 2012

    - by cj
    This posts highlights some great scripting language sessions coming up at the Oracle OpenWorld and MySQL Connect conferences. These events are happening in San Francisco from the end of September. You can search for other interesting conference sessions in the Content Catalog. Also check out what is happening at JavaOne in that event's Content Catalog (I haven't included sessions from it in this post.) To find the timeslots and locations of each session, click their respective link and check the "Session Schedule" box on the top right. GEN8431 - General Session: What’s New in Oracle Database Application Development This general session takes a look at what’s been new in the last year in Oracle Database application development tools using the latest generation of database technology. Topics range from Oracle SQL Developer and Oracle Application Express to Java and PHP. (Thomas Kyte - Architect, Oracle) BOF9858 - Meet the Developers of Database Access Services (OCI, ODBC, DRCP, PHP, Python) This session is your opportunity to meet in person the Oracle developers who have built Oracle Database access tools and products such as the Oracle Call Interface (OCI), Oracle C++ Call Interface (OCCI), and Open Database Connectivity (ODBC) drivers; Transparent Application Failover (TAF); Oracle Database Instant Client; Database Resident Connection Pool (DRCP); Oracle Net Services, and so on. The team also works with those who develop the PHP, Ruby, Python, and Perl adapters for Oracle Database. Come discuss with them the features you like, your pains, and new product enhancements in the latest database technology. CON8506 - Syndication and Consolidation: Oracle Database Driver for MySQL Applications This technical session presents a new Oracle Database driver that enables you to run MySQL applications (written in PHP, Perl, C, C++, and so on) against Oracle Database with almost no code change. Use cases for such a driver include application syndication such as interoperability across a relationship database management system, application migration, and database consolidation. In addition, the session covers enhancements in database technology that enable and simplify the migration of third-party databases and applications to and consolidation with Oracle Database. Attend this session to learn more and see a live demo. (Srinath Krishnaswamy - Director, Software Development, Oracle. Kuassi Mensah - Director Product Management, Oracle. Mohammad Lari - Principal Technical Staff, Oracle ) CON9167 - Current State of PHP and MySQL Together, PHP and MySQL power large parts of the Web. The developers of both technologies continue to enhance their software to ensure that developers can be satisfied despite all their changing and growing needs. This session presents an overview of changes in PHP 5.4, which was released earlier this year and shows you various new MySQL-related features available for PHP, from transparent client-side caching to direct support for scaling and high-availability needs. (Johannes Schlüter - SoftwareDeveloper, Oracle) CON8983 - Sharding with PHP and MySQL In deploying MySQL, scale-out techniques can be used to scale out reads, but for scaling out writes, other techniques have to be used. To distribute writes over a cluster, it is necessary to shard the database and store the shards on separate servers. This session provides a brief introduction to traditional MySQL scale-out techniques in preparation for a discussion on the different sharding techniques that can be used with MySQL server and how they can be implemented with PHP. You will learn about static and dynamic sharding schemes, their advantages and drawbacks, techniques for locating and moving shards, and techniques for resharding. (Mats Kindahl - Senior Principal Software Developer, Oracle) CON9268 - Developing Python Applications with MySQL Utilities and MySQL Connector/Python This session discusses MySQL Connector/Python and the MySQL Utilities component of MySQL Workbench and explains how to write MySQL applications in Python. It includes in-depth explanations of the features of MySQL Connector/Python and the MySQL Utilities library, along with example code to illustrate the concepts. Those interested in learning how to expand or build their own utilities and connector features will benefit from the tips and tricks from the experts. This session also provides an opportunity to meet directly with the engineers and provide feedback on your issues and priorities. You can learn what exists today and influence future developments. (Geert Vanderkelen - Software Developer, Oracle) BOF9141 - MySQL Utilities and MySQL Connector/Python: Python Developers, Unite! Come to this lively discussion of the MySQL Utilities component of MySQL Workbench and MySQL Connector/Python. It includes in-depth explanations of the features and dives into the code for those interested in learning how to expand or build their own utilities and connector features. This is an audience-driven session, so put on your best Python shirt and let’s talk about MySQL Utilities and MySQL Connector/Python. (Geert Vanderkelen - Software Developer, Oracle. Charles Bell - Senior Software Developer, Oracle) CON3290 - Integrating Oracle Database with a Social Network Facebook, Flickr, YouTube, Google Maps. There are many social network sites, each with their own APIs for sharing data with them. Most developers do not realize that Oracle Database has base tools for communicating with these sites, enabling all manner of information, including multimedia, to be passed back and forth between the sites. This technical presentation goes through the methods in PL/SQL for connecting to, and then sending and retrieving, all types of data between these sites. (Marcelle Kratochvil - CTO, Piction) CON3291 - Storing and Tuning Unstructured Data and Multimedia in Oracle Database Database administrators need to learn new skills and techniques when the decision is made in their organization to let Oracle Database manage its unstructured data. They will face new scalability challenges. A single row in a table can become larger than a whole database. This presentation covers the techniques a DBA needs for managing the large volume of data in a standard Oracle Database instance. (Marcelle Kratochvil - CTO, Piction) CON3292 - Using PHP, Perl, Visual Basic, Ruby, and Python for Multimedia in Oracle Database These five programming languages are just some of the most popular ones in use at the moment in the marketplace. This presentation details how you can use them to access and retrieve multimedia from Oracle Database. It covers programming techniques and methods for achieving faster development against Oracle Database. (Marcelle Kratochvil - CTO, Piction) UGF5181 - Building Real-World Oracle DBA Tools in Perl Perl is not normally associated with building mission-critical application or DBA tools. Learn why Perl could be a good choice for building your next killer DBA app. This session draws on real-world experience of building DBA tools in Perl, showing the framework and architecture needed to deal with portability, efficiency, and maintainability. Topics include Perl frameworks; Which Comprehensive Perl Archive Network (CPAN) modules are good to use; Perl and CPAN module licensing; Perl and Oracle connectivity; Compiling and deploying your app; An example of what is possible with Perl. (Arjen Visser - CEO & CTO, Dbvisit Software Limited) CON3153 - Perl: A DBA’s and Developer’s Best (Forgotten) Friend This session reintroduces Perl as a language of choice for many solutions for DBAs and developers. Discover what makes Perl so successful and why it is so versatile in our day-to-day lives. Perl can automate all those manual tasks and is truly platform-independent. Perl may not be in the limelight the way other languages are, but it is a remarkable language, it is still very current with ongoing development, and it has amazing online resources. Learn what makes Perl so great (including CPAN), get an introduction to Perl language syntax, find out what you can use Perl for, hear how Oracle uses Perl, discover the best way to learn Perl, and take away a small Perl project challenge. (Arjen Visser - CEO & CTO, Dbvisit Software Limited) CON10332 - Oracle RightNow CX Cloud Service’s Connect PHP API: Intro, What’s New, and Roadmap Connect PHP is a public API that enables developers to build solutions with the Oracle RightNow CX Cloud Service platform. This API is used primarily by developers working within the Oracle RightNow Customer Portal Cloud Service framework who are looking to gain access to data and services hosted by the Oracle RightNow CX Cloud Service platform through a backward-compatible API. Connect for PHP leverages the same data model and services as the Connect Web Services for SOAP API. Come to this session to get an introduction and learn what’s new and what’s coming up. (Mark Rhoads - Senior Principal Applications Engineer, Oracle. Mark Ericson - Sr. Principle Product Manager, Oracle) CON10330 - Oracle RightNow CX Cloud Service APIs and Frameworks Overview Oracle RightNow CX Cloud Service APIs are available in the following areas: desktop UI, Web services, customer portal, PHP, and knowledge. These frameworks provide access to Oracle RightNow CX Cloud Service’s Connect Common Object Model and custom objects. This session provides a broad overview of capabilities in all these areas. (Mark Ericson - Sr. Principle Product Manager, Oracle)

    Read the article

  • Creating an SMF service for mercurial web server

    - by Chris W Beal
    I'm working on a project at the moment, which has a number of contributers. We're managing the project gate (which is stand alone) with mercurial. We want to have an easy way of seeing the changelog, so we can show management what is going on.  Luckily mercurial provides a basic web server which allows you to see the changes, and drill in to change sets. This can be run as a daemon, but as it was running on our build server, every time it was rebooted, someone needed to remember to start the process again. This is of course a classic usage of SMF. Now I'm not an experienced person at writing SMF services, so it took me 1/2 an hour or so to figure it out the first time. But going forward I should know what I'm doing a bit better. I did reference this doc extensively. Taking a step back, the command to start the mercurial web server is $ hg serve -p <port number> -d So we somehow need to get SMF to run that command for us. In the simplest form, SMF services are really made up of two components. The manifest Usually lives in /var/svc/manifest somewhere Can be imported from any location The method Usually live in /lib/svc/method I simply put the script straight in that directory. Not very repeatable, but it worked Can take an argument of start, stop, or refresh Lets start with the manifest. This looks pretty complex, but all it's doing is describing the service name, the dependencies, the start and stop methods, and some properties. The properties can be by instance, that is to say I could have multiple hg serve processes handling different mercurial projects, on different ports simultaneously Here is the manifest I wrote. I stole extensively from the examples in the Documentation. So my manifest looks like this $ cat hg-serve.xml <?xml version="1.0"?> <!DOCTYPE service_bundle SYSTEM "/usr/share/lib/xml/dtd/service_bundle.dtd.1"> <service_bundle type='manifest' name='hg-serve'> <service name='application/network/hg-serve' type='service' version='1'> <dependency name='network' grouping='require_all' restart_on='none' type='service'> <service_fmri value='svc:/milestone/network:default' /> </dependency> <exec_method type='method' name='start' exec='/lib/svc/method/hg-serve %m' timeout_seconds='2' /> <exec_method type='method' name='stop' exec=':kill' timeout_seconds='2'> </exec_method> <instance name='project-gate' enabled='true'> <method_context> <method_credential user='root' group='root' /> </method_context> <property_group name='hg-serve' type='application'> <propval name='path' type='astring' value='/src/project-gate'/> <propval name='port' type='astring' value='9998' /> </property_group> </instance> <stability value='Evolving' /> <template> <common_name> <loctext xml:lang='C'>hg-serve</loctext> </common_name> <documentation> <manpage title='hg' section='1' /> </documentation> </template> </service> </service_bundle> So the only things I had to decide on in this are the service name "application/network/hg-serve" the start and stop methods (more of which later) and the properties. This is the information I need to pass to the start method script. In my case the port I want to start the web server on "9998", and the path to the source gate "/src/project-gate". These can be read in to the start method. So now lets look at the method scripts $ cat /lib/svc/method/hg-serve #!/sbin/sh # # # Copyright (c) 2012, Oracle and/or its affiliates. All rights reserved. # # Standard prolog # . /lib/svc/share/smf_include.sh if [ -z $SMF_FMRI ]; then echo "SMF framework variables are not initialized." exit $SMF_EXIT_ERR fi # # Build the command line flags # # Get the port and directory from the SMF properties port=`svcprop -c -p hg-serve/port $SMF_FMRI` dir=`svcprop -c -p hg-serve/path $SMF_FMRI` echo "$1" case "$1" in 'start') cd $dir /usr/bin/hg serve -d -p $port ;; *) echo "Usage: $0 {start|refresh|stop}" exit 1 ;; esac exit $SMF_EXIT_OK This is all pretty self explanatory, we read the port and directory using svcprop, and use those simply to run a command in the start case. We don't need to implement a stop case, as the manifest says to use "exec=':kill'for the stop method. Now all we need to do is import the manifest and start the service, but first verify the manifest # svccfg verify /path/to/hg-serve.xml If that doesn't give an error try importing it # svccfg import /path/to/hg-serve.xml If like me you originally put the hg-serve.xml file in /var/svc/manifest somewhere you'll get an error and told to restart the import service svccfg: Restarting svc:/system/manifest-import The manifest being imported is from a standard location and should be imported with the command : svcadm restart svc:/system/manifest-import # svcadm restart svc:/system/manifest-import and you're nearly done. You can look at the service using svcs -l # svcs -l hg-serve fmri svc:/application/network/hg-serve:project-gate name hg-serve enabled false state disabled next_state none state_time Thu May 31 16:11:47 2012 logfile /var/svc/log/application-network-hg-serve:project-gate.log restarter svc:/system/svc/restarter:default contract_id 15749 manifest /var/svc/manifest/network/hg/hg-serve.xml dependency require_all/none svc:/milestone/network:default (online) And look at the interesting properties # svcprop hg-serve hg-serve/path astring /src/project-gate hg-serve/port astring 9998 ...stuff deleted.... Then simply enable the service and if every things gone right, you can point your browser at http://server:9998 and get a nice graphical log of project activity. # svcadm enable hg-serve # svcs -l hg-serve fmri svc:/application/network/hg-serve:project-gate name hg-serve enabled true state online next_state none state_time Thu May 31 16:18:11 2012 logfile /var/svc/log/application-network-hg-serve:project-gate.log restarter svc:/system/svc/restarter:default contract_id 15858 manifest /var/svc/manifest/network/hg/hg-serve.xml dependency require_all/none svc:/milestone/network:default (online) None of this is rocket science, but a bit fiddly. Hence I thought I'd blog it. It might just be you see this in google and it clicks with you more than one of the many other blogs or how tos about it. Plus I can always refer back to it myself in 3 weeks, when I want to add another project to the server, and I've forgotten how to do it.

    Read the article

  • Oracle Tutor: Top 10 to Implement Sustainable Policies and Procedures

    - by emily.chorba(at)oracle.com
    Overview Your organization (executives, managers, and employees) understands the value of having written business process documents (process maps, procedures, instructions, reference documents, and form abstracts). Policies and procedures should be documented because they help to reduce the range of individual decisions and encourage management by exception: the manager only needs to give special attention to unusual problems, not covered by a specific policy or procedure. As more and more procedures are written to cover recurring situations, managers will begin to make decisions which will be consistent from one functional area to the next.Companies should take a project management approach when implementing an environment for a sustainable documentation program and do the following:1. Identify an Executive Champion2. Put together a winning team3. Assign ownership4. Centralize publishing5. Establish the Document Maintenance Process Up Front6. Document critical activities only7. Document actual practice8. Minimize documentation9. Support continuous improvement10. Keep it simple 1. Identify an Executive ChampionAppoint a top down driver. Select one key individual to be a mentor for the procedure planning team. The individual should be a senior manager, such as your company president, CIO, CFO, the vice-president of quality, manufacturing, or engineering. Written policies and procedures can be important supportive aids when known to express the thinking for the chief executive officer and / or the president and to have his or her full support. 2. Put Together a Winning TeamChoose a strong Project Management Leader and staff the procedure planning team with management members from cross functional groups. Make sure team members have the responsibility - and the authority - to make things happen.The winning team should consist of the Documentation Project Manager, Document Owners (one for each functional area), a Document Controller, and Document Specialists (as needed). The Tutor Implementation Guide has complete job descriptions for these roles. 3. Assign Ownership It is virtually impossible to keep process documentation simple and meaningful if employees who are far removed from the activity itself create it. It is impossible to keep documentation up-to-date when responsibility for the document is not clearly understood.Key to the Tutor methodology, therefore, is the concept of ownership. Each document has a single owner, who is responsible for ensuring that the document is necessary and that it reflects actual practice. The owner must be a person who is knowledgeable about the activity and who has the authority to build consensus among the persons who participate in the activity as well as the authority to define or change the way an activity is performed. The owner must be an advocate of the performers and negotiate, not dictate practices.In the Tutor environment, a document's owner is the only person with the authority to approve an update to that document. 4. Centralize Publishing Although it is tempting (especially in a networked environment and with document management software solutions) to decentralize the control of all documents -- with each owner updating and distributing his own -- Tutor promotes centralized publishing by assigning the Document Administrator (gate keeper) to manage the updates and distribution of the procedures library. 5. Establish a Document Maintenance Process Up Front (and stick to it) Everyone in your organization should know they are invited to suggest changes to procedures and should understand exactly what steps to take to do so. Tutor provides a set of procedures to help your company set up a healthy document control system. There are many document management products available to automate some of the document change and maintenance steps. Depending on the size of your organization, a simple document management system can reduce the effort it takes to track and distribute document changes and updates. Whether your company decides to store the written policies and procedures on a file server or in a database, the essential tasks for maintaining documents are the same, though some tasks are automated. 6. Document Critical Activities Only The best way to keep your documentation simple is to reduce the number of process documents to a bare minimum and to include in those documents only as much detail as is absolutely necessary. The first step to reducing process documentation is to document only those activities that are deemed critical. Not all activities require documentation. In fact, some critical activities cannot and should not be standardized. Others may be sufficiently documented with an instruction or a checklist and may not require a procedure. A document should only be created when it enhances the performance of the employee performing the activity. If it does not help the employee, then there is no reason to maintain the document. Activities that represent little risk (such as project status), activities that cannot be defined in terms of specific tasks (such as product research), and activities that can be performed in a variety of ways (such as advertising) often do not require documentation. Sometimes, an activity will evolve to the point where documentation is necessary. For example, an activity performed by single employee may be straightforward and uncomplicated -- that is, until the activity is performed by multiple employees. Sometimes, it is the interaction between co-workers that necessitates documentation; sometimes, it is the complexity or the diversity of the activity.7. Document Actual Practices The only reason to maintain process documentation is to enhance the performance of the employee performing the activity. And documentation can only enhance performance if it reflects reality -- that is, current best practice. Documentation that reflects an unattainable ideal or outdated practices will end up on the shelf, unused and forgotten.Documenting actual practice means (1) auditing the activity to understand how the work is really performed, (2) identifying best practices with employees who are involved in the activity, (3) building consensus so that everyone agrees on a common method, and (4) recording that consensus.8. Minimize Documentation One way to keep it simple is to document at the highest level possible. That is, include in your documents only as much detail as is absolutely necessary.When writing a document, you should ask yourself, What is the purpose of this document? That is, what problem will it solve?By focusing on this question, you can target the critical information.• What questions are the end users likely to have?• What level of detail is required?• Is any of this information extraneous to the document's purpose? Short, concise documents are user friendly and they are easier to keep up to date. 9. Support Continuous Improvement Employees who perform an activity are often in the best position to identify improvements to the process. In other words, continuous improvement is a natural byproduct of the work itself -- but only if the improvements are communicated to all employees who are involved in the process, and only if there is consensus among those employees.Traditionally, process documentation has been used to dictate performance, to limit employees' actions. In the Tutor environment, process documents are used to communicate improvements identified by employees. How does this work? The Tutor methodology requires a process document to reflect actual practice, so the owner of a document must routinely audit its content -- does the document match what the employees are doing? If it doesn't, the owner has the responsibility to evaluate the process, to build consensus among the employees, to identify "best practices," and to communicate these improvements via a document update. Continuous improvement can also be an outgrowth of corrective action -- but only if the solutions to problems are communicated effectively. The goal should be to solve a problem once and only once, which means not only identifying the solution, but ensuring that the solution becomes part of the process. The Tutor system provides the method through which improvements and solutions are documented and communicated to all affected employees in a cost-effective, timely manner; it ensures that improvements are not lost or confined to a single employee. 10. Keep it Simple Process documents don't have to be complex and unfriendly. In fact, the simpler the format and organization, the more likely the documents will be used. And the simpler the method of maintenance, the more likely the documents will be kept up-to-date. Keep it simply by:• Minimizing skills and training required• Following the established Tutor document format and layout• Avoiding technology just for technology's sake No other rule has as major an impact on the success of your internal documentation as -- keep it simple. Learn More For more information about Tutor, visit Oracle.Com or the Tutor Blog. Post your questions at the Tutor Forum.   Emily Chorba Principle Product Manager Oracle Tutor & BPM 

    Read the article

  • Full-text Indexing Books Online

    - by Most Valuable Yak (Rob Volk)
    While preparing for a recent SQL Saturday presentation, I was struck by a crazy idea (shocking, I know): Could someone import the content of SQL Server Books Online into a database and apply full-text indexing to it?  The answer is yes, and it's really quite easy to do. The first step is finding the installed help files.  If you have SQL Server 2012, BOL is installed under the Microsoft Help Library.  You can find the install location by opening SQL Server Books Online and clicking the gear icon for the Help Library Manager.  When the new window pops up click the Settings link, you'll get the following: You'll see the path under Library Location. Once you navigate to that path you'll have to drill down a little further, to C:\ProgramData\Microsoft\HelpLibrary\content\Microsoft\store.  This is where the help file content is kept if you downloaded it for offline use. Depending on which products you've downloaded help for, you may see a few hundred files.  Fortunately they're named well and you can easily find the "SQL_Server_Denali_Books_Online_" files.  We are interested in the .MSHC files only, and can skip the Installation and Developer Reference files. Despite the .MHSC extension, these files are compressed with the standard Zip format, so your favorite archive utility (WinZip, 7Zip, WinRar, etc.) can open them.  When you do, you'll see a few thousand files in the archive.  We are only interested in the .htm files, but there's no harm in extracting all of them to a folder.  7zip provides a command-line utility and the following will extract to a D:\SQLHelp folder previously created: 7z e –oD:\SQLHelp "C:\ProgramData\Microsoft\HelpLibrary\content\Microsoft\store\SQL_Server_Denali_Books_Online_B780_SQL_110_en-us_1.2.mshc" *.htm Well that's great Rob, but how do I put all those files into a full-text index? I'll tell you in a second, but first we have to set up a few things on the database side.  I'll be using a database named Explore (you can certainly change that) and the following setup is a fragment of the script I used in my presentation: USE Explore; GO CREATE SCHEMA help AUTHORIZATION dbo; GO -- Create default fulltext catalog for later FT indexes CREATE FULLTEXT CATALOG FTC AS DEFAULT; GO CREATE TABLE help.files(file_id int not null IDENTITY(1,1) CONSTRAINT PK_help_files PRIMARY KEY, path varchar(256) not null CONSTRAINT UNQ_help_files_path UNIQUE, doc_type varchar(6) DEFAULT('.xml'), content varbinary(max) not null); CREATE FULLTEXT INDEX ON help.files(content TYPE COLUMN doc_type LANGUAGE 1033) KEY INDEX PK_help_files; This will give you a table, default full-text catalog, and full-text index on that table for the content you're going to insert.  I'll be using the command line again for this, it's the easiest method I know: for %a in (D:\SQLHelp\*.htm) do sqlcmd -S. -E -d Explore -Q"set nocount on;insert help.files(path,content) select '%a', cast(c as varbinary(max)) from openrowset(bulk '%a', SINGLE_CLOB) as c(c)" You'll need to copy and run that as one line in a command prompt.  I'll explain what this does while you run it and watch several thousand files get imported: The "for" command allows you to loop over a collection of items.  In this case we want all the .htm files in the D:\SQLHelp folder.  For each file it finds, it will assign the full path and file name to the %a variable.  In the "do" clause, we'll specify another command to be run for each iteration of the loop.  I make a call to "sqlcmd" in order to run a SQL statement.  I pass in the name of the server (-S.), where "." represents the local default instance. I specify -d Explore as the database, and -E for trusted connection.  I then use -Q to run a query that I enclose in double quotes. The query uses OPENROWSET(BULK…SINGLE_CLOB) to open the file as a data source, and to treat it as a single character large object.  In order for full-text indexing to work properly, I have to convert the text content to varbinary. I then INSERT these contents along with the full path of the file into the help.files table created earlier.  This process continues for each file in the folder, creating one new row in the table. And that's it! 5 SQL Statements and 2 command line statements to unzip and import SQL Server Books Online!  In case you're wondering why I didn't use FILESTREAM or FILETABLE, it's simply because I haven't learned them…yet. I may return to this blog after I figure that out and update it with the steps to do so.  I believe that will make it even easier. In the spirit of exploration, I'll leave you to work on some fulltext queries of this content.  I also recommend playing around with the sys.dm_fts_xxxx DMVs (I particularly like sys.dm_fts_index_keywords, it's pretty interesting).  There are additional example queries in the download material for my presentation linked above. Many thanks to Kevin Boles (t) for his advice on (re)checking the content of the help files.  Don't let that .htm extension fool you! The 2012 help files are actually XML, and you'd need to specify '.xml' in your document type column in order to extract the full-text keywords.  (You probably noticed this in the default definition for the doc_type column.)  You can query sys.fulltext_document_types to get a complete list of the types that can be full-text indexed. I also need to thank Hilary Cotter for giving me the original idea. I believe he used MSDN content in a full-text index for an article from waaaaaaaaaaay back, that I can't find now, and had forgotten about until just a few days ago.  He is also co-author of Pro Full-Text Search in SQL Server 2008, which I highly recommend.  He also has some FTS articles on Simple Talk: http://www.simple-talk.com/sql/learn-sql-server/sql-server-full-text-search-language-features/ http://www.simple-talk.com/sql/learn-sql-server/sql-server-full-text-search-language-features,-part-2/

    Read the article

  • The Presentation Isn't Over Until It's Over

    - by Phil Factor
    The senior corporate dignitaries settled into their seats looking important in a blue-suited sort of way. The lights dimmed as I strode out in front to give my presentation.  I had ten vital minutes to make my pitch.  I was about to dazzle the top management of a large software company who were considering the purchase of my software product. I would present them with a dazzling synthesis of diagrams, graphs, followed by  a live demonstration of my software projected from my laptop.  My preparation had been meticulous: It had to be: A year’s hard work was at stake, so I’d prepared it to perfection.  I stood up and took them all in, with a gaze of sublime confidence. Then the laptop expired. There are several possible alternative plans of action when this happens     A. Stare at the smoking laptop vacuously, flapping ones mouth slowly up and down     B. Stand frozen like a statue, locked in indecision between fright and flight.     C. Run out of the room, weeping     D. Pretend that this was all planned     E. Abandon the presentation in favour of a stilted and tedious dissertation about the software     F. Shake your fist at the sky, and curse the sense of humour of your preferred deity I started for a few seconds on plan B, normally referred to as the ‘Rabbit in the headlamps of the car’ technique. Suddenly, a little voice inside my head spoke. It spoke the famous inane words of Yogi Berra; ‘The game isn't over until it's over.’ ‘Too right’, I thought. What to do? I ran through the alternatives A-F inclusive in my mind but none appealed to me. I was completely unprepared for this. Nowadays, longevity has since taught me more than I wanted to know about the wacky sense of humour of fate, and I would have taken two laptops. I hadn’t, but decided to do the presentation anyway as planned. I started out ignoring the dead laptop, but pretending, instead that it was still working. The audience looked startled. They were expecting plan B to be succeeded by plan C, I suspect. They weren’t used to denial on this scale. After my introductory talk, which didn’t require any visuals, I came to the diagram that described the application I’d written.  I’d taken ages over it and it was hot stuff. Well, it would have been had it been projected onto the screen. It wasn’t. Before I describe what happened then, I must explain that I have thespian tendencies.  My  triumph as Professor Higgins in My Fair Lady at the local operatic society is now long forgotten, but I remember at the time of my finest performance, the moment that, glancing up over the vast audience of  moist-eyed faces at the during the poignant  scene between Eliza and Higgins at the end, I  realised that I had a talent that one day could possibly  be harnessed for commercial use I just talked about the diagram as if it was there, but throwing in some extra description. The audience nodded helpfully when I’d done enough. Emboldened, I began a sort of mime, well, more of a ballet, to represent each slide as I came to it. Heaven knows I’d done my preparation and, in my mind’s eye, I could see every detail, but I had to somehow project the reality of that vision to the audience, much the same way any actor playing Macbeth should do the ghost of Banquo.  My desperation gave me a manic energy. If you’ve ever demonstrated a windows application entirely by mime, gesture and florid description, you’ll understand the scale of the challenge, but then I had nothing to lose. With a brief sentence of description here and there, and arms flailing whilst outlining the size and shape of  graphs and diagrams, I used the many tricks of mime, gesture and body-language  learned from playing Captain Hook, or the Sheriff of Nottingham in pantomime. I set out determinedly on my desperate venture. There wasn’t time to do anything but focus on the challenge of the task: the world around me narrowed down to ten faces and my presentation: ten souls who had to be hypnotized into seeing a Windows application:  one that was slick, well organized and functional I don’t remember the details. Eight minutes of my life are gone completely. I was a thespian berserker.  I know however that I followed the basic plan of building the presentation in a carefully controlled crescendo until the dazzling finale where the results were displayed on-screen.  ‘And here you see the results, neatly formatted and grouped carefully to enhance the significance of the figures, together with running trend-graphs!’ I waved a mime to signify an animated  window-opening, and looked up, in my first pause, to gaze defiantly  at the audience.  It was a sight I’ll never forget. Ten pairs of eyes were gazing in rapt attention at the imaginary window, and several pairs of eyes were glancing at the imaginary graphs and figures.  I hadn’t had an audience like that since my starring role in  Beauty and the Beast.  At that moment, I realized that my desperate ploy might work. I sat down, slightly winded, when my ten minutes were up.  For the first and last time in my life, the audience of a  ‘PowerPoint’ presentation burst into spontaneous applause. ‘Any questions?’ ‘Yes,  Have you got an agent?’ Yes, in case you’re wondering, I got the deal. They bought the software product from me there and then. However, it was a life-changing experience for me and I have never ever again trusted technology as part of a presentation.  Even if things can’t go wrong, they’ll go wrong and they’ll kill the flow of what you’re presenting.  if you can’t do something without the techno-props, then you shouldn’t do it.  The greatest lesson of all is that great presentations require preparation and  ‘stage-presence’ rather than fancy graphics. They’re a great supporting aid, but they should never dominate to the point that you’re lost without them.

    Read the article

  • ApiChange Is Released!

    - by Alois Kraus
    I have been working on little tool to simplify my life and perhaps yours as developer as well. It is basically a command line tool that allows you to execute queries on your compiled .NET code base. The main purpose is to find out how big the impact of an api change would be if you changed this or that.  Now you can do high level operations like Diff public types for breaking changes. Who uses a method? Who uses a type? Who uses implements an interface? Who references me? What format has the binary  (32/64, Managed C++, Pure IL, Unmanaged)? Search for all event subscribers and unsubscribers. A unique feature is to check for event subscription imbalances. Forgotten event subscriptions are the 90% cause of managed memory leaks. It is done at a per class level. If one class does subscribe to one event more often than it does unsubscribe it is treated as possible event subscription imbalance. Another unique ability is to search for users of string literals which allows you to track users of a string constant which is not possible otherwise. For incremental builds the ShowRebuildTargets command can be used to identify the dependant targets that need a rebuild after you did compile one assembly. It has some heuristics in place to determine the impact of breaking changes and finds out which targets need to be recompiled as well. It has a ton of other features and a an API to access these things programmatically so you can build upon these simple queries create even better tools. Perhaps we get a Visual Studio plug in? You can download it from CodePlex here. It works via XCopy deployment. Simply let it run and check the command line help out. The best feature in my opinion is that the output of nearly all commands can be piped to Excel for further analysis. Since it does read also the pdbs it can show you the source file name and line number as well for all matches. The following picture shows the output of a –WhousesType query. The following command checks where type from BaseLibraryV1.dll are used inside DependantLibV1.dll. All matches are printed out with the reason and matching item along with file and line number. There is even a hyper link to the match which will open Visual Studio. ApiChange -whousestype "*" BaseLibraryV1.dll -in DependantLibV1.dll –excel The "*” is the actual query which means all types. The syntax is the same like in C# just that placeholders are allowed ;-). More info's can be found at the Codeplex Documentation.     The tool was developed in a TDD style manner which means that it is heavily tested and already used by a quite large user base inside the company I do work for. Luckily for you I got the permission to make it public so you take advantage of it. It is fully instrumented with tracing. If you find bugs simply add the –trace command line switch to find out what is failing and send me the output. How is it done? Your first guess might be that it uses reflection. Wrong. It is based on Mono Cecil a free IL parser with a fantastic API to access all internals of a managed assembly. The speed is awesome and to make it even faster I did make the tool heavily multi threaded. The query above did execute in 1.8s with the Excel output. On a rather slow machine I can analyze over 1500 assemblies in less than 40s with a very low memory consumption. The true power of Mono Cecil is that I can load an assembly like any other data file. I have no problems unloading a file but if I would have used reflection I would need to unload a whole AppDomain just to get rid of one assembly in my memory. Just to give you a glimpse how ApiChange.Api.dll can be used I show you one of the unit tests:           public void Can_Find_GenericMethodInvocations_With_Type_Parameters()         { // 1. Create an aggregator to collect our matches             UsageQueryAggregator agg = new UsageQueryAggregator();   // 2. This is the type we want to search for. Load it via the type query             var decimalType = TypeQuery.GetTypeByName(TestConstants.MscorlibAssembly, "System.Decimal");   // 3. register the type query which searches for uses of the Decimal type             new WhoUsesType(agg, decimalType);   // 4. Search for all users of the Decimal type in the DependandLibV1Assembly             agg.Analyze(TestConstants.DependandLibV1Assembly);   // Extract matches and assert             Assert.AreEqual(2, agg.MethodMatches.Count, "Method match count");             Assert.AreEqual("UseGenericMethod", agg.MethodMatches[0].Match.Name);             Assert.AreEqual("UseGenericMethod", agg.MethodMatches[1].Match.Name);         } Many thanks go from here to Jb Evian for the creation of Mono.Cecil. Without this fantastic piece of code it would have been much much harder. There are other options around like the Common Compiler Infrastructure  Metadata Api which should do the same thing but it was not a real option since the Microsoft reader did fail on even simple assemblies (at least in September 2009 this was the case). Besides this I found the CCI Apis much harder to use. The only real competitor was Reflector which does support many things but does not let me access his cool high level analyze commands. So I decided to dig into the IL specs and as a result you can query your compiled binaries from the command line or programmatically. The best thing is you try it out for yourself and give me some feedback what you miss. If you want to contribute or have a cool idea what should be added drop me a mail at A Kraus1@___No [email protected]. There is much more inside the tool I did not talk about it (yet).

    Read the article

  • 5 Lessons learnt in localization / multi language support in WPF

    - by MarkPearl
    For the last few months I have been secretly working away at the second version of an application that we initially released a few years ago. It’s called MaxCut and it is a free panel/cut optimizer for the woodwork, glass and metal industry. One of the motivations for writing MaxCut was to get an end to end experience in developing an application for general consumption. From the early days of v1 of MaxCut I would get the odd email thanking me for the software and then listing a few suggestions on how to improve it. Two of the most dominant suggestions that we received were… Support for imperial measurements (the original program only supported the metric system) Multi language support (we had someone who volunteered to translate the program into Japanese for us). I am not going to dive into the Imperial to Metric support in todays blog post, but I would like to cover a few brief lessons we learned in adding support for multi-language functionality in the software. I have sectioned them below under different lessons. Lesson 1 – Build multi-language support in from the start So the first lesson I learnt was if you know you are going to do multi language support – build it in from the very beginning! One of the power points of WPF/Silverlight is data binding in XAML and so while it wasn’t to painful to retro fit multi language support into the programing, it was still time consuming and a bit tedious to go through mounds and mounds of views and would have been a minor job to have implemented this while the form was being designed. Lesson 2 – Accommodate for varying word lengths using Grids The next lesson was a little harder to learn and was learnt a bit further down the road in the development cycle. We developed everything in English, assuming that other languages would have similar character length words for equivalent meanings… don’t!. A word that is short in your language may be of varying character lengths in other languages. Some language like Dutch and German allow for concatenation of nouns which has the potential to create really long words. We picked up a few places where our views had been structured incorrectly so that if a word was to long it would get clipped off or cut out. To get around this we began using the WPF grid extensively with column widths that would automatically expand if they needed to. Generally speaking the grid replacement got round this hurdle, and if in future you have a choice between a stack panel or a grid – think twice before going for the easier option… often the grid will be a bit more work to setup, but will be more flexible. Lesson 3 – Separate the separators Our initial run through moving the words to a resource dictionary led us to make what I thought was one potential mistake. If we had a label like the following… “length : “ In the resource dictionary we put it as a single entry. This is fine until you start using a word more than once. For instance in our scenario we used the word “length’ frequently. with different variations of the word with grammar and separators included in the resource we ended up having what I would consider a bloated dictionary. When we removed the separators from the words and put them as their own resources we saw a dramatic reduction in dictionary size… so something that looked like this… “length : “ “length. “ “length?” Was reduced to… “length” “:” “?” “.” While this may not seem like a reduction at first glance, consider that the separators “:?.” are used everywhere and suddenly you see a real reduction in bloat. Lesson 4 – Centralize the Language Dictionary This lesson was learnt at the very end of the project after we had already had a release candidate out in the wild. Because our translations would be done on a volunteer basis and remotely, we wanted it to be really simple for someone to translate our program into another language. As a common design practice we had tiered the application so that we had a business logic layer, a ui layer, etc. The problem was in several of these layers we had resource files specific for that layer. What this resulted in was us having multiple resource files that we would need to send to our translators. To add to our problems, some of the wordings were duplicated in different resource files, which would result in additional frustration from our translators as they felt they were duplicating work. Eventually the workaround was to make a separate project in VS2010 with just the language translations. We then exposed the dictionary as public within this project and made it as a reference to the other projects within the solution. This solved out problem as now we had a central dictionary and could remove any duplication's. Lesson 5 – Make a dummy translation file to test that you haven’t missed anything The final lesson learnt about multi language support in WPF was when checking if you had forgotten to translate anything in the inline code, make a test resource file with dummy data. Ideally you want the data for each word to be identical. In our instance we made one which had all the resource key values pointing to a value of test. This allowed us point the language file to our test resource file and very quickly browse through the program and see if we had missed any linking. The alternative to this approach is to have two language files and swap between the two while running the program to make sure that you haven’t missed anything, but the downside of dual language file approach is that it is much a lot harder spotting a mistake if everything is different – almost like playing Where’s Wally / Waldo. It is much easier spotting variance in uniformity – meaning when you put the “test’ keyword for everything, anything that didn’t say “test” stuck out like a sore thumb. So these are my top five lessons learnt on implementing multi language support in WPF. Feel free to make any suggestions in the comments section if you feel maybe something is more important than one of these or if I got it wrong!

    Read the article

  • Metrics - A little knowledge can be a dangerous thing (or 'Why you're not clever enough to interpret metrics data')

    - by Jason Crease
    At RedGate Software, I work on a .NET obfuscator  called SmartAssembly.  Various features of it use a database to store various things (exception reports, name-mappings, etc.) The user is given the option of using either a SQL-Server database (which requires them to have Microsoft SQL Server), or a Microsoft Access MDB file (which requires nothing). MDB is the default option, but power-users soon switch to using a SQL Server database because it offers better performance and data-sharing. In the fashionable spirit of optimization and metrics, an obvious product-management question is 'Which is the most popular? SQL Server or MDB?' We've collected data about this fact, using our 'Feature-Usage-Reporting' technology (available as part of SmartAssembly) and more recently our 'Application Metrics' technology: Parameter Number of users % of total users Number of sessions Number of usages SQL Server 28 19.0 8115 8115 MDB 114 77.6 1449 1449 (As a disclaimer, please note than SmartAssembly has far more than 132 users . This data is just a selection of one build) So, it would appear that SQL-Server is used by fewer users, but more often. Great. But here's why these numbers are useless to me: Only the original developers understand the data What does a single 'usage' of 'MDB' mean? Does this happen once per run? Once per option change? On clicking the 'Obfuscate Now' button? When running the command-line version or just from the UI version? Each question could skew the data 10-fold either way, and the answers only known by the developer that instrumented the application in the first place. In other words, only the original developer can interpret the data - product-managers cannot interpret the data unaided. Most of the data is from uninterested users About half of people who download and run a free-trial from the internet quit it almost immediately. Only a small fraction use it sufficiently to make informed choices. Since the MDB option is the default one, we don't know how many of those 114 were people CHOOSING to use the MDB, or how many were JUST HAPPENING to use this MDB default for their 20-second trial. This is a problem we see across all our metrics: Are people are using X because it's the default or are they using X because they want to use X? We need to segment the data further - asking what percentage of each percentage meet our criteria for an 'established user' or 'informed user'. You end up spending hours writing sophisticated and dubious SQL queries to segment the data further. Not fun. You can't find out why they used this feature Metrics can answer the when and what, but not the why. Why did people use feature X? If you're anything like me, you often click on random buttons in unfamiliar applications just to explore the feature-set. If we listened uncritically to metrics at RedGate, we would eliminate the most-important and more-complex features which people actually buy the software for, leaving just big buttons on the main page and the About-Box. "Ah, that's interesting!" rather than "Ah, that's actionable!" People do love data. Did you know you eat 1201 chickens in a lifetime? But just 4 cows? Interesting, but useless. Often metrics give you a nice number: '5.8% of users have 3 or more monitors' . But unless the statistic is both SUPRISING and ACTIONABLE, it's useless. Most metrics are collected, reviewed with lots of cooing. and then forgotten. Unless a piece-of-data could change things, it's useless collecting it. People get obsessed with significance levels The first things that lots of people do with this data is do a t-test to get a significance level ("Hey! We know with 99.64% confidence that people prefer SQL Server to MDBs!") Believe me: other causes of error/misinterpretation in your data are FAR more significant than your t-test could ever comprehend. Confirmation bias prevents objectivity If the data appears to match our instinct, we feel satisfied and move on. If it doesn't, we suspect the data and dig deeper, plummeting down a rabbit-hole of segmentation and filtering until we give-up and move-on. Data is only useful if it can change our preconceptions. Do you trust this dodgy data more than your own understanding, knowledge and intelligence?  I don't. There's always multiple plausible ways to interpret/action any data Let's say we segment the above data, and get this data: Post-trial users (i.e. those using a paid version after the 14-day free-trial is over): Parameter Number of users % of total users Number of sessions Number of usages SQL Server 13 9.0 1115 1115 MDB 5 4.2 449 449 Trial users: Parameter Number of users % of total users Number of sessions Number of usages SQL Server 15 10.0 7000 7000 MDB 114 77.6 1000 1000 How do you interpret this data? It's one of: Mostly SQL Server users buy our software. People who can't afford SQL Server tend to be unable to afford or unwilling to buy our software. Therefore, ditch MDB-support. Our MDB support is so poor and buggy that our massive MDB user-base doesn't buy it.  Therefore, spend loads of money improving it, and think about ditching SQL-Server support. People 'graduate' naturally from MDB to SQL Server as they use the software more. Things are fine the way they are. We're marketing the tool wrong. The large number of MDB users represent uninformed downloaders. Tell marketing to aggressively target SQL Server users. To choose an interpretation you need to segment again. And again. And again, and again. Opting-out is correlated with feature-usage Metrics tends to be opt-in. This skews the data even further. Between 5% and 30% of people choose to opt-in to metrics (often called 'customer improvement program' or something like that). Casual trial-users who are uninterested in your product or company are less likely to opt-in. This group is probably also likely to be MDB users. How much does this skew your data by? Who knows? It's not all doom and gloom. There are some things metrics can answer well. Environment facts. How many people have 3 monitors? Have Windows 7? Have .NET 4 installed? Have Japanese Windows? Minor optimizations.  Is the text-box big enough for average user-input? Performance data. How long does our app take to start? How many databases does the average user have on their server? As you can see, questions about who-the-user-is rather than what-the-user-does are easier to answer and action. Conclusion Use SmartAssembly. If not for the metrics (called 'Feature-Usage-Reporting'), then at least for the obfuscation/error-reporting. Data raises more questions than it answers. Questions about environment are the easiest to answer.

    Read the article

  • What's up with LDoms: Part 5 - A few Words about Consoles

    - by Stefan Hinker
    Back again to look at a detail of LDom configuration that is often forgotten - the virtual console server. Remember, LDoms are SPARC systems.  As such, each guest will have it's own OBP running.  And to connect to that OBP, the administrator will need a console connection.  Since it's OBP, and not some x86 BIOS, this console will be very serial in nature ;-)  It's really very much like in the good old days, where we had a terminal concentrator where all those serial cables ended up in.  Just like with other components in LDoms, the virtualized solution looks very similar. Every LDom guest requires exactly one console connection.  Envision this similar to the RS-232 port on older SPARC systems.  The LDom framework provides one or more console services that provide access to these connections.  This would be the virtual equivalent of a network terminal server (NTS), where all those serial cables are plugged in.  In the physical world, we'd have a list somewhere, that would tell us which TCP-Port of the NTS was connected to which server.  "ldm list" does just that: root@sun # ldm list NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME primary active -n-cv- UART 16 7680M 0.4% 27d 8h 22m jupiter bound ------ 5002 20 8G mars active -n---- 5000 2 8G 0.5% 55d 14h 10m venus active -n---- 5001 2 8G 0.5% 56d 40m pluto inactive ------ 4 4G The column marked "CONS" tells us, where to reach the console of each domain. In the case of the primary domain, this is actually a (more) physical connection - it's the console connection of the physical system, which is either reachable via the ILOM of that system, or directly via the serial console port on the chassis. All the other guests are reachable through the console service which we created during the inital setup of the system.  Note that pluto does not have a port assigned.  This is because pluto is not yet bound.  (Binding can be viewed very much as the assembly of computer parts - CPU, Memory, disks, network adapters and a serial console cable are all put together when binding the domain.)  Unless we set the port number explicitly, LDoms Manager will do this on a first come, first serve basis.  For just a few domains, this is fine.  For larger deployments, it might be a good idea to assign these port numbers manually using the "ldm set-vcons" command.  However, there is even better magic associated with virtual consoles. You can group several domains into one console group, reachable through one TCP port of the console service.  This can be useful when several groups of administrators are to be given access to different domains, or for other grouping reasons.  Here's an example: root@sun # ldm set-vcons group=planets service=console jupiter root@sun # ldm set-vcons group=planets service=console pluto root@sun # ldm bind jupiter root@sun # ldm bind pluto root@sun # ldm list NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME primary active -n-cv- UART 16 7680M 6.1% 27d 8h 24m jupiter bound ------ 5002 200 8G mars active -n---- 5000 2 8G 0.6% 55d 14h 12m pluto bound ------ 5002 4 4G venus active -n---- 5001 2 8G 0.5% 56d 42m root@sun # telnet localhost 5002 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. sun-vnts-planets: h, l, c{id}, n{name}, q:l DOMAIN ID DOMAIN NAME DOMAIN STATE 2 jupiter online 3 pluto online sun-vnts-planets: h, l, c{id}, n{name}, q:npluto Connecting to console "pluto" in group "planets" .... Press ~? for control options .. What I did here was add the two domains pluto and jupiter to a new console group called "planets" on the service "console" running in the primary domain.  Simply using a group name will create such a group, if it doesn't already exist.  By default, each domain has its own group, using the domain name as the group name.  The group will be available on port 5002, chosen by LDoms Manager because I didn't specify it.  If I connect to that console group, I will now first be prompted to choose the domain I want to connect to from a little menu. Finally, here's an example how to assign port numbers explicitly: root@sun # ldm set-vcons port=5044 group=pluto service=console pluto root@sun # ldm bind pluto root@sun # ldm list NAME STATE FLAGS CONS VCPU MEMORY UTIL UPTIME primary active -n-cv- UART 16 7680M 3.8% 27d 8h 54m jupiter active -t---- 5002 200 8G 0.5% 30m mars active -n---- 5000 2 8G 0.6% 55d 14h 43m pluto bound ------ 5044 4 4G venus active -n---- 5001 2 8G 0.4% 56d 1h 13m With this, pluto would always be reachable on port 5044 in its own exclusive console group, no matter in which order other domains are bound. Now, you might be wondering why we always have to mention the console service name, "console" in all the examples here.  The simple answer is because there could be more than one such console service.  For all "normal" use, a single console service is absolutely sufficient.  But the system is flexible enough to allow more than that single one, should you need them.  In fact, you could even configure such a console service on a domain other than the primary (or control domain), which would make that domain a real console server.  I actually have a customer who does just that - they want to separate console access from the control domain functionality.  But this is definately a rather sophisticated setup. Something I don't want to go into in this post is access control.  vntsd, which is the daemon providing all these console services, is fully RBAC-aware, and you can configure authorizations for individual users to connect to console groups or individual domain's consoles.  If you can't wait until I get around to security, check out the man page of vntsd. Further reading: The Admin Guide is rather reserved on this subject.  I do recommend to check out the Reference Manual. The manpage for vntsd will discuss all the control sequences as well as the grouping and authorizations mentioned here.

    Read the article

  • Option Trading: Getting the most out of the event session options

    - by extended_events
    You can control different aspects of how an event session behaves by setting the event session options as part of the CREATE EVENT SESSION DDL. The default settings for the event session options are designed to handle most of the common event collection situations so I generally recommend that you just use the defaults. Like everything in the real world though, there are going to be a handful of “special cases” that require something different. This post focuses on identifying the special cases and the correct use of the options to accommodate those cases. There is a reason it’s called Default The default session options specify a total event buffer size of 4 MB with a 30 second latency. Translating this into human terms; this means that our default behavior is that the system will start processing events from the event buffer when we reach about 1.3 MB of events or after 30 seconds, which ever comes first. Aside: What’s up with the 1.3 MB, I thought you said the buffer was 4 MB?The Extended Events engine takes the total buffer size specified by MAX_MEMORY (4MB by default) and divides it into 3 equally sized buffers. This is done so that a session can be publishing events to one buffer while other buffers are being processed. There are always at least three buffers; how to get more than three is covered later. Using this configuration, the Extended Events engine can “keep up” with most event sessions on standard workloads. Why is this? The fact is that most events are small, really small; on the order of a couple hundred bytes. Even when you start considering events that carry dynamically sized data (eg. binary, text, etc.) or adding actions that collect additional data, the total size of the event is still likely to be pretty small. This means that each buffer can likely hold thousands of events before it has to be processed. When the event buffers are finally processed there is an economy of scale achieved since most targets support bulk processing of the events so they are processed at the buffer level rather than the individual event level. When all this is working together it’s more likely that a full buffer will be processed and put back into the ready queue before the remaining buffers (remember, there are at least three) are full. I know what you’re going to say: “My server is exceptional! My workload is so massive it defies categorization!” OK, maybe you weren’t going to say that exactly, but you were probably thinking it. The point is that there are situations that won’t be covered by the Default, but that’s a good place to start and this post assumes you’ve started there so that you have something to look at in order to determine if you do have a special case that needs different settings. So let’s get to the special cases… What event just fired?! How about now?! Now?! If you believe the commercial adage from Heinz Ketchup (Heinz Slow Good Ketchup ad on You Tube), some things are worth the wait. This is not a belief held by most DBAs, particularly DBAs who are looking for an answer to a troubleshooting question fast. If you’re one of these anxious DBAs, or maybe just a Program Manager doing a demo, then 30 seconds might be longer than you’re comfortable waiting. If you find yourself in this situation then consider changing the MAX_DISPATCH_LATENCY option for your event session. This option will force the event buffers to be processed based on your time schedule. This option only makes sense for the asynchronous targets since those are the ones where we allow events to build up in the event buffer – if you’re using one of the synchronous targets this option isn’t relevant. Avoid forgotten events by increasing your memory Have you ever had one of those days where you keep forgetting things? That can happen in Extended Events too; we call it dropped events. In order to optimizes for server performance and help ensure that the Extended Events doesn’t block the server if to drop events that can’t be published to a buffer because the buffer is full. You can determine if events are being dropped from a session by querying the dm_xe_sessions DMV and looking at the dropped_event_count field. Aside: Should you care if you’re dropping events?Maybe not – think about why you’re collecting data in the first place and whether you’re really going to miss a few dropped events. For example, if you’re collecting query duration stats over thousands of executions of a query it won’t make a huge difference to miss a couple executions. Use your best judgment. If you find that your session is dropping events it means that the event buffer is not large enough to handle the volume of events that are being published. There are two ways to address this problem. First, you could collect fewer events – examine you session to see if you are over collecting. Do you need all the actions you’ve specified? Could you apply a predicate to be more specific about when you fire the event? Assuming the session is defined correctly, the next option is to change the MAX_MEMORY option to a larger number. Picking the right event buffer size might take some trial and error, but a good place to start is with the number of dropped events compared to the number you’ve collected. Aside: There are three different behaviors for dropping events that you specify using the EVENT_RETENTION_MODE option. The default is to allow single event loss and you should stick with this setting since it is the best choice for keeping the impact on server performance low.You’ll be tempted to use the setting to not lose any events (NO_EVENT_LOSS) – resist this urge since it can result in blocking on the server. If you’re worried that you’re losing events you should be increasing your event buffer memory as described in this section. Some events are too big to fail A less common reason for dropping an event is when an event is so large that it can’t fit into the event buffer. Even though most events are going to be small, you might find a condition that occasionally generates a very large event. You can determine if your session is dropping large events by looking at the dm_xe_sessions DMV once again, this time check the largest_event_dropped_size. If this value is larger than the size of your event buffer [remember, the size of your event buffer, by default, is max_memory / 3] then you need a large event buffer. To specify a large event buffer you set the MAX_EVENT_SIZE option to a value large enough to fit the largest event dropped based on data from the DMV. When you set this option the Extended Events engine will create two buffers of this size to accommodate these large events. As an added bonus (no extra charge) the large event buffer will also be used to store normal events in the cases where the normal event buffers are all full and waiting to be processed. (Note: This is just a side-effect, not the intended use. If you’re dropping many normal events then you should increase your normal event buffer size.) Partitioning: moving your events to a sub-division Earlier I alluded to the fact that you can configure your event session to use more than the standard three event buffers – this is called partitioning and is controlled by the MEMORY_PARTITION_MODE option. The result of setting this option is fairly easy to explain, but knowing when to use it is a bit more art than science. First the science… You can configure partitioning in three ways: None, Per NUMA Node & Per CPU. This specifies the location where sets of event buffers are created with fairly obvious implication. There are rules we follow for sub-dividing the total memory (specified by MAX_MEMORY) between all the event buffers that are specific to the mode used: None: 3 buffers (fixed)Node: 3 * number_of_nodesCPU: 2.5 * number_of_cpus Here are some examples of what this means for different Node/CPU counts: Configuration None Node CPU 2 CPUs, 1 Node 3 buffers 3 buffers 5 buffers 6 CPUs, 2 Node 3 buffers 6 buffers 15 buffers 40 CPUs, 5 Nodes 3 buffers 15 buffers 100 buffers   Aside: Buffer size on multi-processor computersAs the number of Nodes or CPUs increases, the size of the event buffer gets smaller because the total memory is sub-divided into more pieces. The defaults will hold up to this for a while since each buffer set is holding events only from the Node or CPU that it is associated with, but at some point the buffers will get too small and you’ll either see events being dropped or you’ll get an error when you create your session because you’re below the minimum buffer size. Increase the MAX_MEMORY setting to an appropriate number for the configuration. The most likely reason to start partitioning is going to be related to performance. If you notice that running an event session is impacting the performance of your server beyond a reasonably expected level [Yes, there is a reasonably expected level of work required to collect events.] then partitioning might be an answer. Before you partition you might want to check a few other things: Is your event retention set to NO_EVENT_LOSS and causing blocking? (I told you not to do this.) Consider changing your event loss mode or increasing memory. Are you over collecting and causing more work than necessary? Consider adding predicates to events or removing unnecessary events and actions from your session. Are you writing the file target to the same slow disk that you use for TempDB and your other high activity databases? <kidding> <not really> It’s always worth considering the end to end picture – if you’re writing events to a file you can be impacted by I/O, network; all the usual stuff. Assuming you’ve ruled out the obvious (and not so obvious) issues, there are performance conditions that will be addressed by partitioning. For example, it’s possible to have a successful event session (eg. no dropped events) but still see a performance impact because you have many CPUs all attempting to write to the same free buffer and having to wait in line to finish their work. This is a case where partitioning would relieve the contention between the different CPUs and likely reduce the performance impact cause by the event session. There is no DMV you can check to find these conditions – sorry – that’s where the art comes in. This is  largely a matter of experimentation. On the bright side you probably won’t need to to worry about this level of detail all that often. The performance impact of Extended Events is significantly lower than what you may be used to with SQL Trace. You will likely only care about the impact if you are trying to set up a long running event session that will be part of your everyday workload – sessions used for short term troubleshooting will likely fall into the “reasonably expected impact” category. Hey buddy – I think you forgot something OK, there are two options I didn’t cover: STARTUP_STATE & TRACK_CAUSALITY. If you want your event sessions to start automatically when the server starts, set the STARTUP_STATE option to ON. (Now there is only one option I didn’t cover.) I’m going to leave causality for another post since it’s not really related to session behavior, it’s more about event analysis. - Mike Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Option Trading: Getting the most out of the event session options

    - by extended_events
    You can control different aspects of how an event session behaves by setting the event session options as part of the CREATE EVENT SESSION DDL. The default settings for the event session options are designed to handle most of the common event collection situations so I generally recommend that you just use the defaults. Like everything in the real world though, there are going to be a handful of “special cases” that require something different. This post focuses on identifying the special cases and the correct use of the options to accommodate those cases. There is a reason it’s called Default The default session options specify a total event buffer size of 4 MB with a 30 second latency. Translating this into human terms; this means that our default behavior is that the system will start processing events from the event buffer when we reach about 1.3 MB of events or after 30 seconds, which ever comes first. Aside: What’s up with the 1.3 MB, I thought you said the buffer was 4 MB?The Extended Events engine takes the total buffer size specified by MAX_MEMORY (4MB by default) and divides it into 3 equally sized buffers. This is done so that a session can be publishing events to one buffer while other buffers are being processed. There are always at least three buffers; how to get more than three is covered later. Using this configuration, the Extended Events engine can “keep up” with most event sessions on standard workloads. Why is this? The fact is that most events are small, really small; on the order of a couple hundred bytes. Even when you start considering events that carry dynamically sized data (eg. binary, text, etc.) or adding actions that collect additional data, the total size of the event is still likely to be pretty small. This means that each buffer can likely hold thousands of events before it has to be processed. When the event buffers are finally processed there is an economy of scale achieved since most targets support bulk processing of the events so they are processed at the buffer level rather than the individual event level. When all this is working together it’s more likely that a full buffer will be processed and put back into the ready queue before the remaining buffers (remember, there are at least three) are full. I know what you’re going to say: “My server is exceptional! My workload is so massive it defies categorization!” OK, maybe you weren’t going to say that exactly, but you were probably thinking it. The point is that there are situations that won’t be covered by the Default, but that’s a good place to start and this post assumes you’ve started there so that you have something to look at in order to determine if you do have a special case that needs different settings. So let’s get to the special cases… What event just fired?! How about now?! Now?! If you believe the commercial adage from Heinz Ketchup (Heinz Slow Good Ketchup ad on You Tube), some things are worth the wait. This is not a belief held by most DBAs, particularly DBAs who are looking for an answer to a troubleshooting question fast. If you’re one of these anxious DBAs, or maybe just a Program Manager doing a demo, then 30 seconds might be longer than you’re comfortable waiting. If you find yourself in this situation then consider changing the MAX_DISPATCH_LATENCY option for your event session. This option will force the event buffers to be processed based on your time schedule. This option only makes sense for the asynchronous targets since those are the ones where we allow events to build up in the event buffer – if you’re using one of the synchronous targets this option isn’t relevant. Avoid forgotten events by increasing your memory Have you ever had one of those days where you keep forgetting things? That can happen in Extended Events too; we call it dropped events. In order to optimizes for server performance and help ensure that the Extended Events doesn’t block the server if to drop events that can’t be published to a buffer because the buffer is full. You can determine if events are being dropped from a session by querying the dm_xe_sessions DMV and looking at the dropped_event_count field. Aside: Should you care if you’re dropping events?Maybe not – think about why you’re collecting data in the first place and whether you’re really going to miss a few dropped events. For example, if you’re collecting query duration stats over thousands of executions of a query it won’t make a huge difference to miss a couple executions. Use your best judgment. If you find that your session is dropping events it means that the event buffer is not large enough to handle the volume of events that are being published. There are two ways to address this problem. First, you could collect fewer events – examine you session to see if you are over collecting. Do you need all the actions you’ve specified? Could you apply a predicate to be more specific about when you fire the event? Assuming the session is defined correctly, the next option is to change the MAX_MEMORY option to a larger number. Picking the right event buffer size might take some trial and error, but a good place to start is with the number of dropped events compared to the number you’ve collected. Aside: There are three different behaviors for dropping events that you specify using the EVENT_RETENTION_MODE option. The default is to allow single event loss and you should stick with this setting since it is the best choice for keeping the impact on server performance low.You’ll be tempted to use the setting to not lose any events (NO_EVENT_LOSS) – resist this urge since it can result in blocking on the server. If you’re worried that you’re losing events you should be increasing your event buffer memory as described in this section. Some events are too big to fail A less common reason for dropping an event is when an event is so large that it can’t fit into the event buffer. Even though most events are going to be small, you might find a condition that occasionally generates a very large event. You can determine if your session is dropping large events by looking at the dm_xe_sessions DMV once again, this time check the largest_event_dropped_size. If this value is larger than the size of your event buffer [remember, the size of your event buffer, by default, is max_memory / 3] then you need a large event buffer. To specify a large event buffer you set the MAX_EVENT_SIZE option to a value large enough to fit the largest event dropped based on data from the DMV. When you set this option the Extended Events engine will create two buffers of this size to accommodate these large events. As an added bonus (no extra charge) the large event buffer will also be used to store normal events in the cases where the normal event buffers are all full and waiting to be processed. (Note: This is just a side-effect, not the intended use. If you’re dropping many normal events then you should increase your normal event buffer size.) Partitioning: moving your events to a sub-division Earlier I alluded to the fact that you can configure your event session to use more than the standard three event buffers – this is called partitioning and is controlled by the MEMORY_PARTITION_MODE option. The result of setting this option is fairly easy to explain, but knowing when to use it is a bit more art than science. First the science… You can configure partitioning in three ways: None, Per NUMA Node & Per CPU. This specifies the location where sets of event buffers are created with fairly obvious implication. There are rules we follow for sub-dividing the total memory (specified by MAX_MEMORY) between all the event buffers that are specific to the mode used: None: 3 buffers (fixed)Node: 3 * number_of_nodesCPU: 2.5 * number_of_cpus Here are some examples of what this means for different Node/CPU counts: Configuration None Node CPU 2 CPUs, 1 Node 3 buffers 3 buffers 5 buffers 6 CPUs, 2 Node 3 buffers 6 buffers 15 buffers 40 CPUs, 5 Nodes 3 buffers 15 buffers 100 buffers   Aside: Buffer size on multi-processor computersAs the number of Nodes or CPUs increases, the size of the event buffer gets smaller because the total memory is sub-divided into more pieces. The defaults will hold up to this for a while since each buffer set is holding events only from the Node or CPU that it is associated with, but at some point the buffers will get too small and you’ll either see events being dropped or you’ll get an error when you create your session because you’re below the minimum buffer size. Increase the MAX_MEMORY setting to an appropriate number for the configuration. The most likely reason to start partitioning is going to be related to performance. If you notice that running an event session is impacting the performance of your server beyond a reasonably expected level [Yes, there is a reasonably expected level of work required to collect events.] then partitioning might be an answer. Before you partition you might want to check a few other things: Is your event retention set to NO_EVENT_LOSS and causing blocking? (I told you not to do this.) Consider changing your event loss mode or increasing memory. Are you over collecting and causing more work than necessary? Consider adding predicates to events or removing unnecessary events and actions from your session. Are you writing the file target to the same slow disk that you use for TempDB and your other high activity databases? <kidding> <not really> It’s always worth considering the end to end picture – if you’re writing events to a file you can be impacted by I/O, network; all the usual stuff. Assuming you’ve ruled out the obvious (and not so obvious) issues, there are performance conditions that will be addressed by partitioning. For example, it’s possible to have a successful event session (eg. no dropped events) but still see a performance impact because you have many CPUs all attempting to write to the same free buffer and having to wait in line to finish their work. This is a case where partitioning would relieve the contention between the different CPUs and likely reduce the performance impact cause by the event session. There is no DMV you can check to find these conditions – sorry – that’s where the art comes in. This is  largely a matter of experimentation. On the bright side you probably won’t need to to worry about this level of detail all that often. The performance impact of Extended Events is significantly lower than what you may be used to with SQL Trace. You will likely only care about the impact if you are trying to set up a long running event session that will be part of your everyday workload – sessions used for short term troubleshooting will likely fall into the “reasonably expected impact” category. Hey buddy – I think you forgot something OK, there are two options I didn’t cover: STARTUP_STATE & TRACK_CAUSALITY. If you want your event sessions to start automatically when the server starts, set the STARTUP_STATE option to ON. (Now there is only one option I didn’t cover.) I’m going to leave causality for another post since it’s not really related to session behavior, it’s more about event analysis. - Mike Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • MVC Automatic Menu

    - by Nuri Halperin
    An ex-colleague of mine used to call his SQL script generator "Super-Scriptmatic 2000". It impressed our then boss little, but was fun to say and use. We called every batch job and script "something 2000" from that day on. I'm tempted to call this one Menu-Matic 2000, except it's waaaay past 2000. Oh well. The problem: I'm developing a bunch of stuff in MVC. There's no PM to generate mounds of requirements and there's no Ux Architect to create wireframe. During development, things change. Specifically, actions get renamed, moved from controller x to y etc. Well, as the site grows, it becomes a major pain to keep a static menu up to date, because the links change. The HtmlHelper doesn't live up to it's name and provides little help. How do I keep this growing list of pesky little forgotten actions reigned in? The general plan is: Decorate every action you want as a menu item with a custom attribute Reflect out all menu items into a structure at load time Render the menu using as CSS  friendly <ul><li> HTML. The MvcMenuItemAttribute decorates an action, designating it to be included as a menu item: [AttributeUsage(AttributeTargets.Method, AllowMultiple = true)] public class MvcMenuItemAttribute : Attribute {   public string MenuText { get; set; }   public int Order { get; set; }   public string ParentLink { get; set; }   internal string Controller { get; set; }   internal string Action { get; set; }     #region ctor   public MvcMenuItemAttribute(string menuText) : this(menuText, 0) { } public MvcMenuItemAttribute(string menuText, int order) { MenuText = menuText; Order = order; }       internal string Link { get { return string.Format("/{0}/{1}", Controller, this.Action); } }   internal MvcMenuItemAttribute ParentItem { get; set; } #endregion } The MenuText allows overriding the text displayed on the menu. The Order allows the items to be ordered. The ParentLink allows you to make this item a child of another menu item. An example action could then be decorated thusly: [MvcMenuItem("Tracks", Order = 20, ParentLink = "/Session/Index")] . All pretty straightforward methinks. The challenge with menu hierarchy becomes fairly apparent when you try to render a menu and highlight the "current" item or render a breadcrumb control. Both encounter an  ambiguity if you allow a data source to have more than one menu item with the same URL link. The issue is that there is no great way to tell which link a person click. Using referring URL will fail if a user bookmarked the page. Using some extra query string to disambiguate duplicate URLs essentially changes the links, and also ads a chance of collision with other query parameters. Besides, that smells. The stock ASP.Net sitemap provider simply disallows duplicate URLS. I decided not to, and simply pick the first one encountered as the "current". Although it doesn't solve the issue completely – one might say they wanted the second of the 2 links to be "current"- it allows one to include a link twice (home->deals and products->deals etc), and the logic of deciding "current" is easy enough to explain to the customer. Now that we got that out of the way, let's build the menu data structure: public static List<MvcMenuItemAttribute> ListMenuItems(Assembly assembly) { var result = new List<MvcMenuItemAttribute>(); foreach (var type in assembly.GetTypes()) { if (!type.IsSubclassOf(typeof(Controller))) { continue; } foreach (var method in type.GetMethods()) { var items = method.GetCustomAttributes(typeof(MvcMenuItemAttribute), false) as MvcMenuItemAttribute[]; if (items == null) { continue; } foreach (var item in items) { if (String.IsNullOrEmpty(item.Controller)) { item.Controller = type.Name.Substring(0, type.Name.Length - "Controller".Length); } if (String.IsNullOrEmpty(item.Action)) { item.Action = method.Name; } result.Add(item); } } } return result.OrderBy(i => i.Order).ToList(); } Using reflection, the ListMenuItems method takes an assembly (you will hand it your MVC web assembly) and generates a list of menu items. It digs up all the types, and for each one that is an MVC Controller, digs up the methods. Methods decorated with the MvcMenuItemAttribute get plucked and added to the output list. Again, pretty simple. To make the structure hierarchical, a LINQ expression matches up all the items to their parent: public static void RegisterMenuItems(List<MvcMenuItemAttribute> items) { _MenuItems = items; _MenuItems.ForEach(i => i.ParentItem = items.FirstOrDefault(p => String.Equals(p.Link, i.ParentLink, StringComparison.InvariantCultureIgnoreCase))); } The _MenuItems is simply an internal list to keep things around for later rendering. Finally, to package the menu building for easy consumption: public static void RegisterMenuItems(Type mvcApplicationType) { RegisterMenuItems(ListMenuItems(Assembly.GetAssembly(mvcApplicationType))); } To bring this puppy home, a call in Global.asax.cs Application_Start() registers the menu. Notice the ugliness of reflection is tucked away from the innocent developer. All they have to do is call the RegisterMenuItems() and pass in the type of the application. When you use the new project template, global.asax declares a class public class MvcApplication : HttpApplication and that is why the Register call passes in that type. protected void Application_Start() { AreaRegistration.RegisterAllAreas(); RegisterRoutes(RouteTable.Routes);   MvcMenu.RegisterMenuItems(typeof(MvcApplication)); }   What else is left to do? Oh, right, render! public static void ShowMenu(this TextWriter output) { var writer = new HtmlTextWriter(output);   renderHierarchy(writer, _MenuItems, null); }   public static void ShowBreadCrumb(this TextWriter output, Uri currentUri) { var writer = new HtmlTextWriter(output); string currentLink = "/" + currentUri.GetComponents(UriComponents.Path, UriFormat.Unescaped);   var menuItem = _MenuItems.FirstOrDefault(m => m.Link.Equals(currentLink, StringComparison.CurrentCultureIgnoreCase)); if (menuItem != null) { renderBreadCrumb(writer, _MenuItems, menuItem); } }   private static void renderBreadCrumb(HtmlTextWriter writer, List<MvcMenuItemAttribute> menuItems, MvcMenuItemAttribute current) { if (current == null) { return; } var parent = current.ParentItem; renderBreadCrumb(writer, menuItems, parent); writer.Write(current.MenuText); writer.Write(" / ");   }     static void renderHierarchy(HtmlTextWriter writer, List<MvcMenuItemAttribute> hierarchy, MvcMenuItemAttribute root) { if (!hierarchy.Any(i => i.ParentItem == root)) return;   writer.RenderBeginTag(HtmlTextWriterTag.Ul); foreach (var current in hierarchy.Where(element => element.ParentItem == root).OrderBy(i => i.Order)) { if (ItemFilter == null || ItemFilter(current)) {   writer.RenderBeginTag(HtmlTextWriterTag.Li); writer.AddAttribute(HtmlTextWriterAttribute.Href, current.Link); writer.AddAttribute(HtmlTextWriterAttribute.Alt, current.MenuText); writer.RenderBeginTag(HtmlTextWriterTag.A); writer.WriteEncodedText(current.MenuText); writer.RenderEndTag(); // link renderHierarchy(writer, hierarchy, current); writer.RenderEndTag(); // li } } writer.RenderEndTag(); // ul } The ShowMenu method renders the menu out to the provided TextWriter. In previous posts I've discussed my partiality to using well debugged, time test HtmlTextWriter to render HTML rather than writing out angled brackets by hand. In addition, writing out using the actual writer on the actual stream rather than generating string and byte intermediaries (yes, StringBuilder being no exception) disturbs me. To carry out the rendering of an hierarchical menu, the recursive renderHierarchy() is used. You may notice that an ItemFilter is called before rendering each item. I figured that at some point one might want to exclude certain items from the menu based on security role or context or something. That delegate is the hook for such future feature. To carry out rendering of a breadcrumb recursion is used again, this time simply to unwind the parent hierarchy from the leaf node, then rendering on the return from the recursion rather than as we go along deeper. I guess I was stuck in LISP that day.. recursion is fun though.   Now all that is left is some usage! Open your Site.Master or wherever you'd like to place a menu or breadcrumb, and plant one of these calls: <% MvcMenu.ShowBreadCrumb(this.Writer, Request.Url); %> to show a breadcrumb trail (notice lack of "=" after <% and the semicolon). <% MvcMenu.ShowMenu(Writer); %> to show the menu.   As mentioned before, the HTML output is nested <UL> <LI> tags, which should make it easy to style using abundant CSS to produce anything from static horizontal or vertical to dynamic drop-downs.   This has been quite a fun little implementation and I was pleased that the code size remained low. The main crux was figuring out how to pass parent information from the attribute to the hierarchy builder because attributes have restricted parameter types. Once I settled on that implementation, the rest falls into place quite easily.

    Read the article

  • Can't launch Oneiric x64 instance on Eucalyptus

    - by Bruno Reis
    EDIT: after many hours, I've found out that the problem has nothing to do with Eucalyptus. It looks like the image is buggy. Very, very buggy. More details in the end. I didn't manage to fix it, and I will file a bug. EDIT 2: I managed to fix it, it apparently works. I have a 4-machine cluster running Ubuntu Server Natty (11.04) x64. I've installed "Ubuntu Enterprise Cloud" from the installtion CD (then updated it) on each of these machines. The cloud seems to work fine, I have lots of virtual machines running Natty servers on them. Now I'd like to run Oneiric in a virtual machine, but somehow I can't. I downloaded Oneiric's (x64) image from http://cloud-images.ubuntu.com/oneiric/current/, published it (uec-publish-tarball oneiric-server-cloudimg-amd64.tar.gz oneiric-server-cloudimg-amd64) exactly as I did with Natty, then tried to launch an instance (euca-run-instances -n 1 -k my-key -t m1.small -z my-cloud emi-XXXXXXXX) using Oneiric's image, but the instance is not able to boot. With euca-get-console-output I get the following: [ 0.461269] VFS: Cannot open root device "sda1" or unknown-block(0,0) [ 0.462388] Please append a correct "root=" boot option; here are the available partitions: [ 0.463855] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) [ 0.465331] Pid: 1, comm: swapper Not tainted 3.0.0-13-generic #22-Ubuntu [ 0.466526] Call Trace: [ 0.466989] [<ffffffff815d3ee5>] panic+0x91/0x194 [ 0.467860] [<ffffffff81ad1031>] mount_block_root+0xdc/0x18e [ 0.468891] [<ffffffff81ad126a>] mount_root+0x54/0x59 [ 0.469829] [<ffffffff81ad13dc>] prepare_namespace+0x16d/0x1a7 [ 0.470883] [<ffffffff81ad0d76>] kernel_init+0x140/0x145 [ 0.471837] [<ffffffff815f38e4>] kernel_thread_helper+0x4/0x10 [ 0.472889] [<ffffffff81ad0c36>] ? start_kernel+0x3df/0x3df [ 0.473884] [<ffffffff815f38e0>] ? gs_change+0x13/0x13 The filesystem is labeled "cloudimg-rootfs", inside the image both /etc/fstab and /boot/grub/grub.cfg always refer to the image by the label, everything seems to be correct, yet the kernel says it can't find the root file system. I've spent many hours googling, but nothing came out. I've asked on #ubuntu-server, but nobody knew what to do. I've asked on #eucalyptus but got no answer at all. Any ideas on why this is happening and how to solve it? Thanks EDIT: after many hours, I've found out that the problem has nothing to do with Eucalyptus. It looks like the image is buggy. Very, very buggy. The first problem is that the Kernel in the image is a -generic kernel, while I suppose it should be a -virtual one. I chrooted into the image, removed the -generic packages, replaced it with the -virtual ones. Then I extracted the new kernel (and replaced the original one (-generic) that came with the tarball) because I need it when I publish and launch an image with Eucalyptus. The problem described above was solved. But then, the console started showing this: mount: mount point ext4 does not exist If you check the /etc/fstab file in the image, it says: LABEL=cloudimg-rootfs ext4 defaults 0 1 Damnt, where's my mount point? Note that it is missing /proc as well. Well, when you think it is over, you will notice that your instance will have no network connectivity. Let's check /etc/network/interface: # interfaces(5) file used by ifup(8) and ifdown(8) auto lo iface lo inet loopback Oh my! It is missing eth0... here I stopped. I can't take no more. I give up. Looks like Canonical has just forgotten to properly set up this image. At first, I though: "have I downloaded a server image by mistake?", but no, I double checked. It is really the cloud image, it has even "cloud-init" installed (which is not, by default, on server images). They just forgot to prepare it. I will file a bug (and reference it here once this is done), and hope they fix it soon! EDIT 2: it looks like the network configuration was the last thing missing. I decided to test it with the fixes above, and it booted properly! However, I haven't got the slightest idea if the image is now good to go...

    Read the article

  • Bad disks in ancient server

    - by Joel Coel
    I have a 1998-era Netware 3.12 server that runs everything on our campus: general ledger, purchasing, payroll, student information, grades, you name it. The server has an Adaptec RAID controller with two volumes: RAID 1, 2 17GB scsi disks, Seagate ST318417W RAID 5, 3 4GB scsi disks, 2 Seagate ST34573W and 1 ST34572W. We are currently in the early stages of a project to replace this system, but you don't just jump into a new system like that and so I need to keep this server running until at least November 2011. This week we had not one but two hard drives fail. Thankfully they are from different volumes and we're able to keep running for the moment, but given the close nature of these failures I have serious doubts that I'll be able to avoid catastrophic failure from this server through the November target as is without restoring the RAID redundancy — it'll only take one more drive failure anywhere and I'm completely hosed. We are fortunate enough to have exact match "spares" lying around for both drives, but the spares are in unknown condition. I tried swapping just them in, but the RAID controller isn't smart enough to handle this and it renders the system unbootable. As for the RAID controller itself, there is utility I can get into during POST via a Ctrl-A shortcut, but I can't do much useful from there. To actually manage volumes I must first boot in to Netware, at which point I can use CI/O Array Management Software Version 2.0 to actually look at volume information. I suspect that the normal way to manage things is to boot from a special floppy with the controller software on it, but that floppy is long gone. Going through the options in the RAID software, I think the only supported way to replace a disk in an existing RAID volume is to physically add the disk, boot up and configure it as a "spare" for a volume, force the volume to use the spare to replace an existing down disk (and at this point I'm only guessing) so that the down disk becomes the spare, repair the volume, remove the spare from the volume, and then shut down and remove the disk. Then start all over for the other failed disk. All this amounts to a lot of downtime, assuming I can even make it work and that my spares are any good. As for finding reliable spares, I have no clue where to even begin looking to find a new 4GB scsi drive, or even which exact scsi system I'm looking for, as it's gone through a few different iterations over time. Another option is to migrate this to a virtual machine (hyper-v), but all previous attempts we've made in this area have failed to get very far. When this machine was installed I was just graduating from high school, and so it requires lower level knowledge of netware and dos than I ever developed, or if I did have since forgotten (I'm not exactly a dos neophyte, either). Part of my problem is this is a high-use server, and taking it down for a few days to figure things out isn't gonna fly very well. As for the question, I'm looking for anything that might be helpful in this situation: a recommendation on a place to find good spares from this era, personal experience repairing RAID volumes using a similar controller or building a hyper-v vm from an old netware server, a line on a floppy with better software for the RAID controller, recommendation on a good Novell consultant in Nebraska that would be able to put things right, a whole other option I haven't considered yet, etc. Update: For backups, we have good (recently verified via restore) backups of the data only -- nothing for the software that actually runs things. Update 2: Just a progress report that I currently have a working Netware 3.12 install in VMWare Virtual Server 2.0, thanks largely to the guide I found here: http://cerbulescubogdan.blogspot.com/2010/11/novell-netware-312-on-vmware.html The next steps are preparing empty netware volumes to match the additional volumes on my existing server, taking a dump of everything on the C:\ drive and netware volumes on my existing server, and figuring out from that information what modules need added to netware, installing my licenses (we do still have that disk, if it's any good), and moving data over. I have approval to bring the server down for a week after the first of the year (sadly not before), so, aside from creating empty volumes, the rest of the work will have to wait until then. Final Update (Jan 5, 2011): I was able to get spares working in both raid arrays without data loss this week. Both are now listed by the controller as "FAULT TOLLERANT" (yay!). I was also able to build on the progress from my last update and now have a functional "spare" server in VMWare Server 2.0. The spare can run and use our erp software, but I can't put it into production because I can't (yet) print from that box (and I have no idea why). Even so, this VM will do in a pinch if I have no other choice, and between it and the repaired RAID arrays I'm comfortable pushing on until I can junk the machine in November.

    Read the article

  • Psychology researcher wants to learn new language

    - by user273347
    I'm currently considering R, matlab, or python, but I'm open to other options. Could you help me pick the best language for my needs? Here are the criteria I have in mind (not in order): Simple to learn. I don't really have a lot of free time, so I'm looking for something that isn't extremely complicated and/or difficult to pick up. I know some C, FWIW. Good for statistics/psychometrics. I do a ton of statistics and psychometrics analysis. A lot of it is basic stuff that I can do with SPSS, but I'd like to play around with the more advanced stuff too (bootstrapping, genetic programming, data mining, neural nets, modeling, etc). I'm looking for a language/environment that can help me run my simpler analyses faster and give me more options than a canned stat package like SPSS. If it can even make tables for me, then it'll be perfect. I also do a fair bit of experimental psychology. I use a canned experiment "programming" software (SuperLab) to make most of my experiments, but I want to be able to program executable programs that I can run on any computer and that can compile the data from the experiments in a spreadsheet. I know python has psychopy and pyepl and matlab has psychtoolbox, but I don't know which one is best. If R had something like this, I'd probably be sold on R already. I'm looking for something regularly used in academe and industry. Everybody else here (including myself, so far) uses canned stat and experiment programming software. One of the reasons I'm trying to learn a programming language is so that I can keep up when I move to another lab. Looking forward to your comments and suggestions. Thank you all for your kind and informative replies. I appreciate it. It's still a tough choice because of so many strong arguments for each language. Python - Thinking about it, I've forgotten so much about C already (I don't even remember what to do with an array) that it might be better for me to start from scratch with a simple program that does what it's supposed to do. It looks like it can do most of the things I'll need it to do, though not as cleanly as R and MATLAB. R - I'm really liking what I'm reading about R. The packages are perfect for my statistical work now. Given the purpose of R, I don't think it's suited to building psychological experiments though. To clarify, what I mean is making a program that presents visual and auditory stimuli to my specifications (hundreds of them in a preset and/or randomized sequence) and records the response data gathered from participants. MATLAB - It's awesome that cognitive and neuro folk are recommending MATLAB, because I'm preparing for the big leap from social and personality psychology to cognitive neuro. The problem is the Uni where I work doesn't have MATLAB licenses (and 3750 GBP for a compiler license is not an option for me haha). Octave looks like a good alternative. PsychToolbox is compatible with Octave, thankfully. SQL - Thanks for the tip. I'll explore that option, too. Python will be the least backbreaking and most useful in the short term. R is well suited to my current work. MATLAB is well suited to my prospective work. It's a tough call, but I think I am now equipped to make a more well-informed decision about where to go next. Thanks again!

    Read the article

  • How to include the login form on the Home index page in MVC

    - by Bernard Larouche
    Hi guys I really need your help for this. I am relatively new to programming and I need help to something that could be easy for a experienced programmer. I would like to get the login form that we get for free in an MVC application on the left sidebar of my Home index page instead of the usual Account/Login page. I am facing some problems. First I need a product object to be displayed on my Home Index page as well. What I did is that I added a product object to the LogOnModel that they provide in the AccountModels class and I created a UserControl (partial view) copying the content of the LogOn.aspx view. Now my Home index.aspx as well as my partial view inherits the LogOnModel class. I can see the Login form on my Home Index page as well as my product object BUT the login Form is never empty. The last username and password always appear there. I know I must have forgotten something or have done something wrong or the way did it is completely wrong !! Please could you give me some advice Thks <%@ Control Language="C#" Inherits="System.Web.Mvc.ViewUserControl<CoderForTradersSite.Models.LogOnModel>" %> <h4>Login Form</h4> <p> Please enter your username and password. <%= Html.ActionLink("Register", "Register") %> if you don't have an account. </p> <% using (Html.BeginForm()) { %> <%= Html.ValidationSummary(true, "Login was unsuccessful. Please correct the errors and try again.") %> <div> <fieldset> <legend>Account Information</legend> <div class="editor-label"> <%= Html.LabelFor(m => m.UserName) %> </div> <div class="editor-field"> <%= Html.TextBoxFor(m => m.UserName) %> <%= Html.ValidationMessageFor(m => m.UserName) %> </div> <div class="editor-label"> <%= Html.LabelFor(m => m.Password) %> </div> <div class="editor-field"> <%= Html.PasswordFor(m => m.Password) %> <%= Html.ValidationMessageFor(m => m.Password) %> </div> <div class="editor-label"> <%= Html.CheckBoxFor(m => m.RememberMe) %> <%= Html.LabelFor(m => m.RememberMe) %> </div> <p> <input type="submit" value="Log On" /> </p> </fieldset> </div> <% } %>

    Read the article

  • C++/msvc6 application crashes due to heap corruption, any hints?

    - by David Alfonso
    Hello all, let me say first that I'm writing this question after months of trying to find out the root of a crash happening in our application. I'll try to detail as much as possible what I've already found out about it. About the application It runs on Windows XP Professional SP2. It's built with Microsoft Visual C++ 6.0 with Service Pack 6. It's MFC based. It uses several external dlls (e.g. Xerces, ZLib or ACE). It has high performance requirements. It does a lot of network and hard disk I/O, but it's also cpu intensive. It has an exception handling mechanism which generates a minidump when an unhandled exception occurs. Facts about the crash It only happens on multiprocessor/multicore machines and under heavy loads of work. It happens at random (neither we nor our client have found a pattern yet). We cannot reproduce the crash on our testing lab. It only happens on some production systems (but always in multicore machines) It always ends up crashing at the same point, although the complete stack is not always the same. Let me add the stack of the crashing thread (obtained using WinDbg, sorry we don't have symbols) ChildEBP RetAddr Args to Child WARNING: Stack unwind information not available. Following frames may be wrong. 030af6c8 7c9206eb 77bfc3c9 01a80000 00224bc3 MyApplication+0x2a85b9 030af960 7c91e9c0 7c92901b 00000ab4 00000000 ntdll!RtlAllocateHeap+0xeac (FPO: [Non-Fpo]) 030af98c 7c9205c8 00000001 00000000 00000000 ntdll!ZwWaitForSingleObject+0xc (FPO: [3,0,0]) 030af9c0 7c920551 01a80898 7c92056d 313adfb0 ntdll!RtlpFreeToHeapLookaside+0x22 (FPO: [2,0,4]) 030afa8c 4ba3ae96 000307da 00130005 00040012 ntdll!RtlFreeHeap+0x1e9 (FPO: [Non-Fpo]) 030afacc 77bfc2e3 0214e384 3087c8d8 02151030 0x4ba3ae96 030afb00 7c91e306 7c80bfc1 00000948 00000001 msvcrt!free+0xc8 (FPO: [Non-Fpo]) 030afb20 0042965b 030afcc0 0214d780 02151218 ntdll!ZwReleaseSemaphore+0xc (FPO: [3,0,0]) 030afb7c 7c9206eb 02e6c471 02ea0000 00000008 MyApplication+0x2965b 030afe60 7c9205c8 02151248 030aff38 7c920551 ntdll!RtlAllocateHeap+0xeac (FPO: [Non-Fpo]) 030afe74 7c92056d 0210bfb8 02151250 02151250 ntdll!RtlpFreeToHeapLookaside+0x22 (FPO: [2,0,4]) 030aff38 77bfc2de 01a80000 00000000 77bfc2e3 ntdll!RtlFreeHeap+0x647 (FPO: [Non-Fpo]) 7c92056d c5ffffff ce7c94be ff7c94be 00ffffff msvcrt!free+0xc3 (FPO: [Non-Fpo]) 7c920575 ff7c94be 00ffffff 12000000 907c94be 0xc5ffffff 7c920579 00ffffff 12000000 907c94be 90909090 0xff7c94be *** WARNING: Unable to verify checksum for xerces-c_2_7.dll *** ERROR: Symbol file could not be found. Defaulted to export symbols for xerces-c_2_7.dll - 7c92057d 12000000 907c94be 90909090 8b55ff8b MyApplication+0xbfffff 7c920581 907c94be 90909090 8b55ff8b 08458bec xerces_c_2_7 7c920585 90909090 8b55ff8b 08458bec 04408b66 0x907c94be 7c920589 8b55ff8b 08458bec 04408b66 0004c25d 0x90909090 7c92058d 08458bec 04408b66 0004c25d 90909090 0x8b55ff8b The address MyApplication+0x2a85b9 corresponds to a call to erase() of a std::list. What I have tried so far Reviewing all the code related to the point where the crash ends happening. Trying to enable pageheap on our testing lab though nothing useful has been found by now. We have substituted the std::list for a C array and then it crashes in other part of the code (although it is related code, it's not in the code where the old list resided). Coincidentally, now it crashes in another erase, though this time of a std::multiset. Let me copy the stack contained in the dump: ntdll.dll!_RtlpCoalesceFreeBlocks@16() + 0x124e bytes ntdll.dll!_RtlFreeHeap@12() + 0x91f bytes msvcrt.dll!_free() + 0xc3 bytes MyApplication.exe!006a4fda() [Frames below may be incorrect and/or missing, no symbols loaded for MyApplication.exe] MyApplication.exe!0069f305() ntdll.dll!_NtFreeVirtualMemory@16() + 0xc bytes ntdll.dll!_RtlpSecMemFreeVirtualMemory@16() + 0x1b bytes ntdll.dll!_ZwWaitForSingleObject@12() + 0xc bytes ntdll.dll!_RtlpFreeToHeapLookaside@8() + 0x26 bytes ntdll.dll!_RtlFreeHeap@12() + 0x114 bytes msvcrt.dll!_free() + 0xc3 bytes c5ffffff() Possible solutions (that I'm aware of) which cannot be applied "Migrate the application to a newer compiler": We are working on this but It's not a solution at the moment. "Enable pageheap (normal or full)": We can't enable pageheap on production machines as this affects performance heavily. I think that's all I remember now, if I have forgotten something I'll add it asap. If you can give me some hint or propose some possible solution, don't hesitate to answer! Thank you in advance for your time and advice.

    Read the article

  • How to get data from struts2 action with jQuery?

    - by Rafa Romero
    I have an Struts2 Web with GoogleMaps. I want to load a list of markers which are saved in a SQL DDBB. To do this, I tried with jQuery and Ajax. Here you are the code: loadMarkers.java public class loadMarkers extends ActionSupport implements ServletRequestAware,ServletResponseAware{ //Variables de sesion/cookie FunctionClass ses; protected HttpServletResponse servletResponse; protected HttpServletRequest servletRequest; private String userID=""; //Variables del marker private List<marker> markersList = new ArrayList<marker>(); public String execute() throws Exception{ FunctionClass friends = new FunctionClass(); //Leemos de la cookie for(Cookie c : servletRequest.getCookies()) { if (c.getName().equals("userID")) userID = (c.getValue()); } System.out.println("en el loadMarkers"); connectionBBDD con = new connectionBBDD(); markersList = con.loadMarkers(userID); return SUCCESS; } I want to use the markerList array in Javascript, to create the markers. This is the struts.xml file: <package name="jsonActions" namespace="/test" extends="json-default"> <action name="LoadMarkers" class="web.localizadroid.maps.loadMarkers"> <interceptor-ref name="basicStack"/> <result type="json"> <param name="root">markersList</param> </result> </action> </package> And here you are the Javascript (jQuery) code: function loadMarkersJ(){ alert("dentro"); $.ajax({ type : "post", url : "LoadMarkers", dataType: "json", success : function(data) { alert(data); var image = new google.maps.MarkerImage ('http://i53.tinypic.com/ettquh.png'); var jSon_Object = eval("(" + data + ")"); //For para analizar los datos (Json) obtenidos de la BBDD for (x = 0; x < jSon_Object.length; x++) { var markersArray = []; var myLatlng = new google.maps.LatLng(jSon_Object[x].lat, jSon_Object[x].lon); markerLoaded = new google.maps.Marker( { position : myLatlng, map : map, icon: image, title: "NOMBRE: " + jSon_Object[x].tarjetName + "\n" + "ANOTACIONES: " + jSon_Object[x].anotaciones + "\n" + "TIME: " + jSon_Object[x].time }); markersArray.push(markerLoaded); if (markersArray) { for (i in markersArray) { alert("entro en forColocaMarkers"); if (markersArray[i].getAnimation() != null) { markersArray[i].setAnimation(null); } else { markersArray[i].setAnimation(google.maps.Animation.BOUNCE); } markersArray[i].setMap(map); } } } } }); } From success : function(data) { to the end, is JavaScript code to create de markers, and this it's OK. The problem is whith ajax, because I don't get to obtain markerList Array by jSon data return...I think that the problem is in the url attribute from $.ajax...I tried loadMarkers.action and loadMarkers, but nothing happens. When I execute the web, only prints the alert alert("dentro"), the alert alert(data) never has been printed. I have forgotten to add the code where I call the Javascript function (loadMarkersJ). Here you are: <p><s:a action="LoadMarkers.action" namespace="/test" onclick="loadMarkersJ(this)">Cargar Marcadores S</s:a></p> Somebody can help me please?

    Read the article

< Previous Page | 12 13 14 15 16 17  | Next Page >