Search Results

Search found 3797 results on 152 pages for 'talk'.

Page 31/152 | < Previous Page | 27 28 29 30 31 32 33 34 35 36 37 38  | Next Page >

  • Database IDs

    - by fatherjack
    Just a quick post, mainly to test out the new blog format but related to a question on the #sqlhelp hashtag. The question came from Justin Dearing (@zippy1981) as: So I take it database_id isn’t an ever incrementing value. #sqlhelp When a new database is created it is given the lowest available ID. This either is in a gap in IDs where a database has been dropped or the database ID is incremented by one from the highest current ID if there are no gaps to fill. To see this in action, connect to your sandbox server and try this: USE MASTER GO CREATE DATABASE cherry GO USE cherry GO SELECT DB_ID() GO CREATE DATABASE grape GO USE grape GO SELECT DB_ID() GO CREATE DATABASE melon GO USE melon GO SELECT DB_ID() GO USE MASTER GO DROP DATABASE grape GO CREATE DATABASE kiwi GO USE kiwi GO SELECT DB_ID() GO USE MASTER GO DROP DATABASE cherry DROP DATABASE melon DROP DATABASE kiwi You should get an incrementing series of database IDs as the databases are created until the last one where the new database gets allocated the ID that is missing because one was dropped.

    Read the article

  • Udacity: Teaching thousands of students to program online using App Engine

    Udacity: Teaching thousands of students to program online using App Engine Join Fred Sauer & Iein Valdez as they talk with Steve Huffman, founder of Reddit and Hipmunk, and Chris Chew, senior software engineer at Udacity. Steve will share his experience teaching a course on web development using App Engine at Udacity, and Chris will talk about his experience building Udacity itself using App Engine. Submit your questions for Steve and Chris to answer live on air. From: GoogleDevelopers Views: 0 0 ratings Time: 00:00 More in Science & Technology

    Read the article

  • Clever memory usage through the years

    - by Ben Emmett
    A friend and I were recently talking about the really clever tricks people have used to get the most out of memory. I thought I’d share my favorites, and would love to hear yours too! Interleaving on drum memory Back in the ye olde days before I’d been born (we’re talking the 50s / 60s here), working memory commonly took the form of rotating magnetic drums. These would spin at a constant speed, and a fixed head would read from memory when the correct part of the drum passed it by, a bit like a primitive platter disk. Because each revolution took a few milliseconds, programmers took to manually arranging information non-sequentially on the drum, timing when an instruction or memory address would need to be accessed, then spacing information accordingly around the edge of the drum, thus reducing the access delay. Similar techniques were still used on hard disks and floppy disks into the 90s, but have become irrelevant with modern disk technologies. The Hashlife algorithm Conway’s Game of Life has attracted numerous implementations over the years, but Bill Gosper’s Hashlife algorithm is particularly impressive. Taking advantage of the repetitive nature of many cellular automata, it uses a quadtree structure to store the hashes of pieces of the overall grid. Over time there are fewer and fewer new structures which need to be evaluated, so it starts to run faster with larger grids, drastically outperforming other algorithms both in terms of speed and the size of grid which can be simulated. The actual amount of memory used is huge, but it’s used in a clever way, so makes the list . Elite’s procedural generation Ok, so this isn’t exactly a memory optimization – more a storage optimization – but it gets an honorable mention anyway. When writing Elite, David Braben and Ian Bell wanted to build a rich world which gamers could explore, but their 22K memory was something of a limitation (for comparison that’s about the size of my avatar picture at the top of this page). They procedurally generated all the characteristics of the 2048 planets in their virtual universe, including the names, which were stitched together using a lookup table of parts of names. In fact the original plans were for 2^52 planets, but it was decided that that was probably too many. Oh, and they did that all in assembly language. Other games of the time used similar techniques too – The Sentinel’s landscape generation algorithm being another example. Modern Garbage Collectors Garbage collection in managed languages like Java and .NET ensures that most of the time, developers stop needing to care about how they use and clean up memory as the garbage collector handles it automatically. Achieving this without killing performance is a near-miraculous feet of software engineering. Much like when learning chemistry, you find that every time you think you understand how the garbage collector works, it turns out to be a mere simplification; that there are yet more complexities and heuristics to help it run efficiently. Of course introducing memory problems is still possible (and there are tools like our memory profiler to help if that happens to you) but they’re much, much rarer. A cautionary note In the examples above, there were good and well understood reasons for the optimizations, but cunningly optimized code has usually had to trade away readability and maintainability to achieve its gains. Trying to optimize memory usage without being pretty confident that there’s actually a problem is doing it wrong. So what have I missed? Tell me about the ingenious (or stupid) tricks you’ve seen people use. Ben

    Read the article

  • Search SSIS packages for table/column references

    - by Nigel Rivett
    A lot of companies now use TFS or some other system and keep all their packages in a single project. This means that a copy of all the packages will end up on your local disk. There is major failing with SSIS that it is sometimes quite difficult to find what a package is actually doing, what it accesses and what it affects. This is a simple dos script which will search through all packages in a folder for a string and write the names of found packages to an output file. Just copy the text to a .bat file (I use aaSearch.bat) in the folder with all the package scripts Change the output filename (twice), change the find string value and run it in a dos window. It works on any text file type so you can also search store procedure scripts – but there are easier ways of doing that. echo. > aaSearch_factSales.txt for /f “delims=” %%a in (‘dir /B /s *.dtsx’) do call :subr “%%a” goto:EOF :subr findstr “factSales” %1 if %ERRORLEVEL% NEQ 1 echo %1 >> aaSearch_factSales.txt goto:EOF

    Read the article

  • Checking who is connected to your server, with PowerShell.

    - by Fatherjack
    There are many occasions when, as a DBA, you want to see who is connected to your SQL Server, along with how they are connecting and what sort of activities they are carrying out. I’m going to look at a couple of ways of getting this information and compare the effort required and the results achieved of each. SQL Server comes with a couple of stored procedures to help with this sort of task – sp_who and its undocumented counterpart sp_who2. There is also the pumped up version of these called sp_whoisactive, written by Adam Machanic which does way more than these procedures. I wholly recommend you try it out if you don’t already know how it works. When it comes to serious interrogation of your SQL Server activity then it is absolutely indispensable. Anyway, back to the point of this blog, we are going to look at getting the information from sp_who2 for a remote server. I wrote this Powershell script a week or so ago and was quietly happy with it for a while. I’m relatively new to Powershell so forgive both my rather low threshold for entertainment and the fact that something so simple is a moderate achievement for me. $Server = 'SERVERNAME' $SMOServer = New-Object Microsoft.SQLServer.Management.SMO.Server $Server # connection and query stuff         $ConnectionStr = "Server=$Server;Database=Master;Integrated Security=True" $Query = "EXEC sp_who2" $Connection = new-object system.Data.SQLClient.SQLConnection $Table = new-object "System.Data.DataTable" $Connection.connectionstring = $ConnectionStr try{ $Connection.open() $Command = $Connection.CreateCommand() $Command.commandtext = $Query $result = $Command.ExecuteReader() $Table.Load($result) } catch{ # Show error $error[0] | format-list -Force } $Title = "Data access processes (" + $Table.Rows.Count + ")" $Table | Out-GridView -Title $Title $Connection.close() So this is pretty straightforward, create an SMO object that represents our chosen server, define a connection to the database and a table object for the results when we get them, execute our query over the connection, load the results into our table object and then, if everything is error free display these results to the PowerShell grid viewer. The query simply gets the results of ‘EXEC sp_who2′ for us. Depending on how many connections there are will influence how long the query runs. The grid viewer lets me sort and search the results so it can be a pretty handy way to locate troublesome connections. Like I say, I was quite pleased with this, it seems a pretty simple script and was working well for me, I have added a few parameters to control the output and give me more specific details but then I see a script that uses the $SMOServer object itself to provide the process information and saves having to define the connection object and query specifications. $Server = 'SERVERNAME' $SMOServer = New-Object Microsoft.SQLServer.Management.SMO.Server $Server $Processes = $SMOServer.EnumProcesses() $Title = "SMO processes (" + $Processes.Rows.Count + ")" $Processes | Out-GridView -Title $Title Create the SMO object of our server and then call the EnumProcesses method to get all the process information from the server. Staggeringly simple! The results are a little different though. Some columns are the same and we can see the same basic information so my first thought was to which runs faster – so that I can get my results more quickly and also so that I place less stress on my server(s). PowerShell comes with a great way of testing this – the Measure-Command function. All you have to do is wrap your piece of code in Measure-Command {[your code here]} and it will spit out the time taken to execute the code. So, I placed both of the above methods of getting SQL Server process connections in two Measure-Command wrappers and pressed F5! The Powershell console goes blank for a while as the code is executed internally when Measure-Command is used but the grid viewer windows appear and the console shows this. You can take the output from Measure-Command and format it for easier reading but in a simple comparison like this we can simply cross refer the TotalMilliseconds values from the two result sets to see how the two methods performed. The query execution method (running EXEC sp_who2 ) is the first set of timings and the SMO EnumProcesses is the second. I have run these on a variety of servers and while the results vary from execution to execution I have never seen the SMO version slower than the other. The difference has varied and the time for both has ranged from sub-second as we see above to almost 5 seconds on other systems. This difference, I would suggest is partly due to the cost overhead of having to construct the data connection and so on where as the SMO EnumProcesses method has the connection to the server already in place and just needs to call back the process information. There is also the difference in the data sets to consider. Let’s take a look at what we get and where the two methods differ Query execution method (sp_who2) SMO EnumProcesses Description - Urn What looks like an XML or JSON representation of the server name and the process ID SPID Spid The process ID Status Status The status of the process Login Login The login name of the user executing the command HostName Host The name of the computer where the  process originated BlkBy BlockingSpid The SPID of a process that is blocking this one DBName Database The database that this process is connected to Command Command The type of command that is executing CPUTime Cpu The CPU activity related to this process DiskIO - The Disk IO activity related to this process LastBatch - The time the last batch was executed from this process. ProgramName Program The application that is facilitating the process connection to the SQL Server. SPID1 - In my experience this is always the same value as SPID. REQUESTID - In my experience this is always 0 - Name In my experience this is always the same value as SPID and so could be seen as analogous to SPID1 from sp_who2 - MemUsage An indication of the memory used by this process but I don’t know what it is measured in (bytes, Kb, Mb…) - IsSystem True or False depending on whether the process is internal to the SQL Server instance or has been created by an external connection requesting data. - ExecutionContextID In my experience this is always 0 so could be analogous to REQUESTID from sp_who2. Please note, these are my own very brief descriptions of these columns, detail can be found from MSDN for columns in the sp_who results here http://msdn.microsoft.com/en-GB/library/ms174313.aspx. Where the columns are common then I would use that description, in other cases then the information returned is purely for interpretation by the reader. Rather annoyingly both result sets have useful information that the other doesn’t. sp_who2 returns Disk IO and LastBatch information which is really useful but the SMO processes method give you IsSystem and MemUsage which have their place in fault diagnosis methods too. So which is better? On reflection I think I prefer to use the sp_who2 method primarily but knowing that the SMO Enumprocesses method is there when I need it is really useful and I’m sure I’ll use it regularly. I’m OK with the fact that it is the slower method because Measure-Command has shown me how close it is to the other option and that it really isn’t a large enough margin to matter.

    Read the article

  • Building Performance Metrics into ASP.NET MVC Applications

    When you're instrumenting an ASP.NET MVC or Web API application to monitor its performance while it is running, it makes sense to use custom performance counters.There are plenty of tools available that read performance counter data, report on it and create alerts based on it. You can then plot application metrics against all sorts of server and workstation metrics.This way, there will always be the right data to guide your tuning efforts.

    Read the article

  • Max Trinidad Sells PowerShell on the Puerto Rican Seashore

    - by SQLBeat
      In this episode, Max Trinidad, Powershell MVP lets me bait him into predicting the future of computing and helps me understand a thing or two about cultural misconceptions around locked men’s restrooms at busy cantinas. We are in beautiful Puerto Rico for this podcast and in honor of that, I try my hand at Espanol. I know as much Spanish as I do BizTalk Server and it shows, embarrassingly so.  Max is always happy but I make him cry on this one and I feel really horrible about it. I promise. It is my function. CLICK BELOW TO LISTEN >>>>>>>CLICK HERE TO LISTEN >>>>>>>>>> CLICK ABOVE TO SHARPEN YOUR CLAYMORE

    Read the article

  • Disaster Recovery Discovery

    - by Rodney Landrum
    Last weekend I joined several of my IT staff on a mission to perform a DR test in our remote CoLo center in a large South East city of the US. Can I be more obtuse? The goal was simple for me as the sole DBA in a throng of Windows, Storage, Network and SAN admins – restore the databases and make them work. There were 4 applications that back ended to 7 SQL Server databases on 4 different SQL Server instances. We would maintain the original server names, but beyond that it was fair game. We had time to prepare so I was able to script out or otherwise automate the recovery process. I used sp_help_revlogin for three of the servers, a bit of a cheat actually because restoring the Master database on the target DR servers was the specified course of action according to the DR procedures ( the caveat “IF REQUIRED” left it open to interpretation. I really wanted to avoid the step of restoring Master for a number of reasons but mainly because I did not want to deal with issues starting SQL Services afterward. Having to account for the location of TempDB and the version conflicts of the resource DBs were just two of the battles I chose not to fight. Not to mention other system database location problems that might arise and prevent SQL from starting.  I was going to have to restore all of the user databases anyway, so I would not really gain any benefit, outside of logins, for taking the time to restore the source Master database over the newly installed one on the fresh server. What I wanted was the ability to restore the Master database as a user database, call it Master_Mine, from a backup on the source system and then use that restored database to script the SQL Logins and passwords on the DR systems. While I did not attempt this on the trip, the thought stuck in my mind and this past week I succeeded at scripting user accounts and passwords using only a restored copy of the Master database. Granted there were several challenges to overcome.  Also, as is usual for any work like this the usual disclaimers apply:  This is not something that I would imagine Microsoft would condone or support and this was really only an experiment for me to learn if it was even possible. While I have tested the process with success, I do not know that I would use this technique in a documented procedure because future updates for SQL Server will render this technique non-functional. I thought at first, incorrectly of course, that I could use sp_help_revlogin on a restored copy of the master database I named Master_Mine.   Since sp_help_revlogin uses system schema objects, sys.syslogins and sys.server_principals, this was not going to work because all results would come from the main Master database. To test this I added a SQL login via SSMS, backed up Master, restored  it as Master_Mine, and then deleted the login.  Even though the test account I created should presumably still be in the Master_Mine database, I should be able to get to it and script out its creation with its password hash so that I would not need to know the password, but any applications that stored that password would not have to be altered in the DR scenario. They would just work as expected. Once I realized that would not work I began looking deeper.  Knowing that sys.syslogins and sys.server_principals are system views, their underlying code should be available with sp_helptext, right? They were. And this led me to discover the two tables sys.sysxlgns and sys.sysprivs, where the data I needed was stored. These tables existed in both the real Master and the restored copy, Master_Mine.  I used this information to tweak the sp_help_revlogin stored procedure to use these tables instead to create the logins cursor used in sp_help_revlogin. For the password hash,  sp_help_revlogin uses the function LoginProperty() which takes a user name and option ‘passwordhash’ to return the hash for the user. Unfortunately, it requires the login to exist in the Master database. This would not work. So another slight modification I had to make was to pull the password hash itself (pwdhash from sys.sysxlgns) into the logins cursor and comment out the section of sp_help_revlogin that uses LoginProperty. Instead, I pass the pwdhash value as the variable @PWD_varbinary to the sp_hexadecimal stored procedure which is also created by and used within the code provided by Microsoft in the link above for sp_help_revlogin. The final challenge: sys.sysxlgns and sys.server_principals are visible only within a Dedicated Administrator Connection (DAC) query window in SSMS or within SQLCDMD.  To open a DAC connection you have to be logged in on the SQL Server itself, via RDP in my case,  and you preface the server name in the query connection with ADMIN:, so that the server connection looks like ADMIN:ServerName. From there you can create the modified stored procedure in the restored copy of a Master database from a source system as whatever name you like, and then run the modified stored procedure. I named my new stored procedure usp_help_revlogin_MyMaster. Upon execution I was happy to see the logins and password hashes that I needed to apply from the source Master database without having to restore over the new Master system database and without the need to access the original server (assuming it was down due to whatever disaster put it in that state). You will note that I am not providing full code samples here of the modifications. I will say that it was a slight bit of work and anyone who needed to do this for whatever reason, could fairly easily roll their own solution with the information provided herein.  My goal, as I said was to prove that this could be done and provide another option if required to ease the burden of getting SQL Servers up and available in an emergency situation where alternatives may be more challenging or otherwise unavailable.  

    Read the article

  • Profiling SharePoint with ANTS Performance Profiler 5.2

    Using ANTS Performance Profiler with SharePoint has, previously, been possible, but not easy. Version 5.2 of ANTS Performance Profiler changes all that, and Chris Allen has put together a straight-forward guide to profiling SharePoint, demonstrating just how much easier it has become.

    Read the article

  • .NET Security Part 4

    - by Simon Cooper
    Finally, in this series, I am going to cover some of the security issues that can trip you up when using sandboxed appdomains. DISCLAIMER: I am not a security expert, and this is by no means an exhaustive list. If you actually are writing security-critical code, then get a proper security audit of your code by a professional. The examples below are just illustrations of the sort of things that can go wrong. 1. AppDomainSetup.ApplicationBase The most obvious one is the issue covered in the MSDN documentation on creating a sandbox, in step 3 – the sandboxed appdomain has the same ApplicationBase as the controlling appdomain. So let’s explore what happens when they are the same, and an exception is thrown. In the sandboxed assembly, Sandboxed.dll (IPlugin is an interface in a partially-trusted assembly, with a single MethodToDoThings on it): public class UntrustedPlugin : MarshalByRefObject, IPlugin { // implements IPlugin.MethodToDoThings() public void MethodToDoThings() { throw new EvilException(); } } [Serializable] internal class EvilException : Exception { public override string ToString() { // show we have read access to C:\Windows // read the first 5 directories Console.WriteLine("Pwned! Mwuahahah!"); foreach (var d in Directory.EnumerateDirectories(@"C:\Windows").Take(5)) { Console.WriteLine(d.FullName); } return base.ToString(); } } And in the controlling assembly: // what can possibly go wrong? AppDomainSetup appDomainSetup = new AppDomainSetup { ApplicationBase = AppDomain.CurrentDomain.SetupInformation.ApplicationBase } // only grant permissions to execute // and to read the application base, nothing else PermissionSet restrictedPerms = new PermissionSet(PermissionState.None); restrictedPerms.AddPermission( new SecurityPermission(SecurityPermissionFlag.Execution)); restrictedPerms.AddPermission( new FileIOPermission(FileIOPermissionAccess.Read, appDomainSetup.ApplicationBase); restrictedPerms.AddPermission( new FileIOPermission(FileIOPermissionAccess.pathDiscovery, appDomainSetup.ApplicationBase); // create the sandbox AppDomain sandbox = AppDomain.CreateDomain("Sandbox", null, appDomainSetup, restrictedPerms); // execute UntrustedPlugin in the sandbox // don't crash the application if the sandbox throws an exception IPlugin o = (IPlugin)sandbox.CreateInstanceFromAndUnwrap("Sandboxed.dll", "UntrustedPlugin"); try { o.MethodToDoThings() } catch (Exception e) { Console.WriteLine(e.ToString()); } And the result? Oops. We’ve allowed a class that should be sandboxed to execute code with fully-trusted permissions! How did this happen? Well, the key is the exact meaning of the ApplicationBase property: The application base directory is where the assembly manager begins probing for assemblies. When EvilException is thrown, it propagates from the sandboxed appdomain into the controlling assembly’s appdomain (as it’s marked as Serializable). When the exception is deserialized, the CLR finds and loads the sandboxed dll into the fully-trusted appdomain. Since the controlling appdomain’s ApplicationBase directory contains the sandboxed assembly, the CLR finds and loads the assembly into a full-trust appdomain, and the evil code is executed. So the problem isn’t exactly that the sandboxed appdomain’s ApplicationBase is the same as the controlling appdomain’s, it’s that the sandboxed dll was in such a place that the controlling appdomain could find it as part of the standard assembly resolution mechanism. The sandbox then forced the assembly to load in the controlling appdomain by throwing a serializable exception that propagated outside the sandbox. The easiest fix for this is to keep the sandbox ApplicationBase well away from the ApplicationBase of the controlling appdomain, and don’t allow the sandbox permissions to access the controlling appdomain’s ApplicationBase directory. If you do this, then the sandboxed assembly can’t be accidentally loaded into the fully-trusted appdomain, and the code can’t be executed. If the plugin does try to induce the controlling appdomain to load an assembly it shouldn’t, a SerializationException will be thrown when it tries to load the assembly to deserialize the exception, and no damage will be done. 2. Loading the sandboxed dll into the application appdomain As an extension of the previous point, you shouldn’t directly reference types or methods in the sandboxed dll from your application code. That loads the assembly into the fully-trusted appdomain, and from there code in the assembly could be executed. Instead, pull out methods you want the sandboxed dll to have into an interface or class in a partially-trusted assembly you control, and execute methods via that instead (similar to the example above with the IPlugin interface). If you need to have a look at the assembly before executing it in the sandbox, either examine the assembly using reflection from within the sandbox, or load the assembly into the Reflection-only context in the application’s appdomain. The code in assemblies in the reflection-only context can’t be executed, it can only be reflected upon, thus protecting your appdomain from malicious code. 3. Incorrectly asserting permissions You should only assert permissions when you are absolutely sure they’re safe. For example, this method allows a caller read-access to any file they call this method with, including your documents, any network shares, the C:\Windows directory, etc: [SecuritySafeCritical] public static string GetFileText(string filePath) { new FileIOPermission(FileIOPermissionAccess.Read, filePath).Assert(); return File.ReadAllText(filePath); } Be careful when asserting permissions, and ensure you’re not providing a loophole sandboxed dlls can use to gain access to things they shouldn’t be able to. Conclusion Hopefully, that’s given you an idea of some of the ways it’s possible to get past the .NET security system. As I said before, this post is not exhaustive, and you certainly shouldn’t base any security-critical applications on the contents of this blog post. What this series should help with is understanding the possibilities of the security system, and what all the security attributes and classes mean and what they are used for, if you were to use the security system in the future.

    Read the article

  • Antenna Aligner Part 3: Kaspersky

    - by Chris George
    Quick one today. Since starting this project, I've been encountering times where Nomad fails to build my app. It would then take repeated attempts at building to then see a build go through successfully. Rob, who works on Nomad at Red Gate, investigated this and it showed that certain parts of the message required to trigger the 'cloud build' were not getting through to the Nomad app, causing the HTTP connection to stall until timeout. After much scratching heads, it turns out that the Kaspersky Internet Security system I have installed on my laptop at home, was being very aggressive and was causing the problem. Perhaps it's trying to protect me from myself? Anyway, we came up with an interim solution why the Nomad guys investigate with Kaspersky by setting Visual Studio to be a trusted application with the Kaspersky settings and setting it to not scan network traffic. Hey presto! This worked and I have not had a single build problem since (other than losing internet connection, or that embarrassing moment when you blame everyone else then realise you've accidentally switched off your wireless on the laptop).

    Read the article

  • DIVs of equal height

    - by Richard Mitchell
    It's the same old old problem you want to make a set of columns the same height but life it too short for the CSS only version. It's technically possible to do but nowadays you can't run the web without having javascript turned on. There must be an easier way. After a short amount of googling I came across a few solutions. A couple were GPL'd which ruled them straight out as I want Red Gate to pay for my mortgage. The best simple solution was found at. http://www.cssnewbie.com/equal-height-columns-with-jquery/...(read more)

    Read the article

  • Database Delivery Patterns and Practices

    Continuous database delivery is an automated process for building, deploying and testing databases to reduce risk and make rapid releases possible. It's enabled by a pipeline that starts when database changes are checked in, and ends when they're deployed to production. The articles collected here will help you understand the theories and methodologies behind every stage of the database delivery pipeline.

    Read the article

  • No Rest for the Virtuous

    - by Chris Massey
    It has been an impressively brutal month in terms of security breaches, and across a whole range of fronts. The "Cablegate" leaks, courtesy of Wikileaks, appear to be in a league of their own. The "Operation Payback" DDoS attacks against PayPal, MasterCard and Visa (not to mention the less successful attack against Amazon) are equally impressive. Even more recently, the Gawker Media Network was subjected to a relatively sophisticated hack attack by Gnosis, with the hackers gaining access to some...(read more)

    Read the article

  • Basic Defensive Database Programming Techniques

    We can all recognise good-quality database code: It doesn't break with every change in the server's configuration, or on upgrade. It isn't affected by concurrent usage, or high workload. In an extract from his forthcoming book, Alex explains just how to go about producing resilient TSQL code that works, and carries on working.

    Read the article

  • Essential Tools for the WPF Novice

    When Michael sets out to do something, there are no half-measures; So when he set out to learn WPF, we all stand to benefit from the thorough research that he put into the task. He wondered what utility applications could assist programming in WPF. Here are the fruits of all his work.

    Read the article

  • The .NET 4.5 async/await Commands in Promise and Practice

    The .NET 4.5 async/await feature provides an opportunity for improving the scalability and performance of applications, particularly where tasks are more effectively done in parallel. The question is: do the scalability gains come at a cost of slowing individual methods? In this article Jon Smith investigates this issue by conducting a side-by-side evaluation of the standard synchronous methods and the new async methods in real applications.

    Read the article

  • Ubuntu KVM Networking

    - by Guest
    I have two VMs ( linux based VM ) running on a UBUNTU/QEMU KVM. I am looking at a simple networking model, where in the VMs have public address and need to talk to the outside world and the outside world should be able to talk to the VMs as well. I did this Created a bridge with A.B.C.D/16 ( class B address ) eth2 is part of the brige with A.B.C.E /32 host address VM's interface ( through TAP ) has A.B.C.F / 16 address eth2 and VM interface are both part of the bridge. Now however, I am unable to ping these address from the outside world. If i remove eth2 from the bridge, I am able to ping the eth2. Please let me know what is going on. Guest

    Read the article

  • Antenna Aligner part 2: Finding the right direction

    - by Chris George
    Last time I managed to get "my first app(tm)" built, published and running on my iPhone. This was really cool, a piece of my code running on my very own device. Ok, so I'm easily pleased! The next challenge was actually trying to determine what it was I wanted this app to do, and how to do it. Reverting back to good old paper and pen, I started sketching out designs for the app. I knew I wanted it to get a list of transmitters, then clicking on a transmitter would display a compass type view, with an arrow pointing the right way. I figured there would not be much point in continuing until I know I could do the graphical part of the project, i.e. the rotating compass, so armed with that reasoning (plus the fact I just wanted to get on and code!), I once again dived into visual studio. Using my friend (google) I found some example code for getting the compass data from the phone using the PhoneGap framework. // onSuccess: Get the current heading // function onSuccess(heading) {    alert('Heading: ' + heading); } navigator.compass.getCurrentHeading(onSuccess, onError); Using the ripple mobile emulator this showed that it was successfully getting the compass heading. But it didn't work when uploaded to my phone. It turns out that the examples I had been looking at were for PhoneGap 1.0, and Nomad uses PhoneGap 1.4.1. In 1.4.1, getCurrentHeading provides a compass object to onSuccess, not just a numeric value, so the code now looks like // onSuccess: Get the current magnetic heading // function onSuccess(heading) {    alert('Heading: ' + heading.magneticHeading); }; navigator.compass.getCurrentHeading(onSuccess, onError); So the lesson learnt from this... read the documentation for the version you are actually using! This does, however, lead to compatibility problems with ripple as it only supports 1.0 which is a real pain. I hope that the ripple system is updated sometime soon.

    Read the article

  • Web.NET event coming in October

    - by Chris Massey
    If you’re a web developer in Europe (or would like an excuse to travel to Europe), you should definitely take a look at the Web.NET event coming in October. It’s being organized by two Italian web maestros (Simone Chiaretta and Ugo Lattanzi) and the session list looks fantastic. The event site pretty much speaks for itself, but here’s a quick version: It’s a free one-day event on October 20th, with a huge variety of great sessions by great speakers, all 100% focused on web development. There’s a pizza-fuelled hackathon in the evening; thrills, spills and hot new skills. It’s a great chance to network with the local (in relative terms) web development community. It’s free (although all donations are very greatly appreciated). It’s in Milan, darling. Here’s what you need to do: Go and register on www.webnetconf.eu, and vote on which sessions you think look the most interesting. I know this will be a difficult process – it’s *very* hard to choose – but persevere! Grab your place when the free tickets become available early next month (places are limited). Come to Milan in October, learn some new skills, meet some great people, and maybe build something awesome if you feel like staying up late. I’ll be there, and hopefully I’ll see you on the day.

    Read the article

  • A Deep Dive into Transport Queues (Part 2)

    Johan Veldhuis completes his 'Deep Dive' by plunging even deeper into the mysteries of MS Exchange's Transport queues that are used to temporarily store messages which are waiting until they are passed through to the next stage, and explains how to change the way they work via configuration settings.

    Read the article

  • Developing Schema Compare for Oracle (Part 2): Dependencies

    - by Simon Cooper
    In developing Schema Compare for Oracle, one of the issues we came across was the size of the databases. As detailed in my last blog post, we had to allow schema pre-filtering due to the number of objects in a standard Oracle database. Unfortunately, this leads to some quite tricky situations regarding object dependencies. This post explains how we deal with these dependencies. 1. Cross-schema dependencies Say, in the following database, you're populating SchemaA, and synchronizing SchemaA.Table1: SOURCE   TARGET CREATE TABLE SchemaA.Table1 ( Col1 NUMBER REFERENCES SchemaB.Table1(Col1));   CREATE TABLE SchemaA.Table1 ( Col1 VARCHAR2(100) REFERENCES SchemaB.Table1(Col1)); CREATE TABLE SchemaB.Table1 ( Col1 NUMBER PRIMARY KEY);   CREATE TABLE SchemaB.Table1 ( Col1 VARCHAR2(100) PRIMARY KEY); We need to do a rebuild of SchemaA.Table1 to change Col1 from a VARCHAR2(100) to a NUMBER. This consists of: Creating a table with the new schema Inserting data from the old table to the new table, with appropriate conversion functions (in this case, TO_NUMBER) Dropping the old table Rename new table to same name as old table Unfortunately, in this situation, the rebuild will fail at step 1, as we're trying to create a NUMBER column with a foreign key reference to a VARCHAR2(100) column. As we're only populating SchemaA, the naive implementation of the object population prefiltering (sticking a WHERE owner = 'SCHEMAA' on all the data dictionary queries) will generate an incorrect sync script. What we actually have to do is: Drop foreign key constraint on SchemaA.Table1 Rebuild SchemaB.Table1 Rebuild SchemaA.Table1, adding the foreign key constraint to the new table This means that in order to generate a correct synchronization script for SchemaA.Table1 we have to know what SchemaB.Table1 is, and that it also needs to be rebuilt to successfully rebuild SchemaA.Table1. SchemaB isn't the schema that the user wants to synchronize, but we still have to load the table and column information for SchemaB.Table1 the same way as any table in SchemaA. Fortunately, Oracle provides (mostly) complete dependency information in the dictionary views. Before we actually read the information on all the tables and columns in the database, we can get dependency information on all the objects that are either pointed at by objects in the schemas we’re populating, or point to objects in the schemas we’re populating (think about what would happen if SchemaB was being explicitly populated instead), with a suitable query on all_constraints (for foreign key relationships) and all_dependencies (for most other types of dependencies eg a function using another function). The extra objects found can then be included in the actual object population, and the sync wizard then has enough information to figure out the right thing to do when we get to actually synchronize the objects. Unfortunately, this isn’t enough. 2. Dependency chains The solution above will only get the immediate dependencies of objects in populated schemas. What if there’s a chain of dependencies? A.tbl1 -> B.tbl1 -> C.tbl1 -> D.tbl1 If we’re only populating SchemaA, the implementation above will only include B.tbl1 in the dependent objects list, whereas we might need to know about C.tbl1 and D.tbl1 as well, in order to ensure a modification on A.tbl1 can succeed. What we actually need is a graph traversal on the dependency graph that all_dependencies represents. Fortunately, we don’t have to read all the database dependency information from the server and run the graph traversal on the client computer, as Oracle provides a method of doing this in SQL – CONNECT BY. So, we can put all the dependencies we want to include together in big bag with UNION ALL, then run a SELECT ... CONNECT BY on it, starting with objects in the schema we’re populating. We should end up with all the objects that might be affected by modifications in the initial schema we’re populating. Good solution? Well, no. For one thing, it’s sloooooow. all_dependencies, on my test databases, has got over 110,000 rows in it, and the entire query, for which Oracle was creating a temporary table to hold the big bag of graph edges, was often taking upwards of two minutes. This is too long, and would only get worse for large databases. But it had some more fundamental problems than just performance. 3. Comparison dependencies Consider the following schema: SOURCE   TARGET CREATE TABLE SchemaA.Table1 ( Col1 NUMBER REFERENCES SchemaB.Table1(col1));   CREATE TABLE SchemaA.Table1 ( Col1 VARCHAR2(100)); CREATE TABLE SchemaB.Table1 ( Col1 NUMBER PRIMARY KEY);   CREATE TABLE SchemaB.Table1 ( Col1 VARCHAR2(100)); What will happen if we used the dependency algorithm above on the source & target database? Well, SchemaA.Table1 has a foreign key reference to SchemaB.Table1, so that will be included in the source database population. On the target, SchemaA.Table1 has no such reference. Therefore SchemaB.Table1 will not be included in the target database population. In the resulting comparison of the two objects models, what you will end up with is: SOURCE  TARGET SchemaA.Table1 -> SchemaA.Table1 SchemaB.Table1 -> (no object exists) When this comparison is synchronized, we will see that SchemaB.Table1 does not exist, so we will try the following sequence of actions: Create SchemaB.Table1 Rebuild SchemaA.Table1, with foreign key to SchemaB.Table1 Oops. Because the dependencies are only followed within a single database, we’ve tried to create an object that already exists. To fix this we can include any objects found as dependencies in the source or target databases in the object population of both databases. SchemaB.Table1 will then be included in the target database population, and we won’t try and create objects that already exist. All good? Well, consider the following schema (again, only explicitly populating SchemaA, and synchronizing SchemaA.Table1): SOURCE   TARGET CREATE TABLE SchemaA.Table1 ( Col1 NUMBER REFERENCES SchemaB.Table1(col1));   CREATE TABLE SchemaA.Table1 ( Col1 VARCHAR2(100)); CREATE TABLE SchemaB.Table1 ( Col1 NUMBER PRIMARY KEY);   CREATE TABLE SchemaB.Table1 ( Col1 VARCHAR2(100) PRIMARY KEY); CREATE TABLE SchemaC.Table1 ( Col1 NUMBER);   CREATE TABLE SchemaC.Table1 ( Col1 VARCHAR2(100) REFERENCES SchemaB.Table1); Although we’re now including SchemaB.Table1 on both sides of the comparison, there’s a third table (SchemaC.Table1) that we don’t know about that will cause the rebuild of SchemaB.Table1 to fail if we try and synchronize SchemaA.Table1. That’s because we’re only running the dependency query on the schemas we’re explicitly populating; to solve this issue, we would have to run the dependency query again, but this time starting the graph traversal from the objects found in the other database. Furthermore, this dependency chain could be arbitrarily extended.This leads us to the following algorithm for finding all the dependencies of a comparison: Find initial dependencies of schemas the user has selected to compare on the source and target Include these objects in both the source and target object populations Run the dependency query on the source, starting with the objects found as dependents on the target, and vice versa Repeat 2 & 3 until no more objects are found For the schema above, this will result in the following sequence of actions: Find initial dependenciesSchemaA.Table1 -> SchemaB.Table1 found on sourceNo objects found on target Include objects in both source and targetSchemaB.Table1 included in source and target Run dependency query, starting with found objectsNo objects to start with on sourceSchemaB.Table1 -> SchemaC.Table1 found on target Include objects in both source and targetSchemaC.Table1 included in source and target Run dependency query on found objectsNo objects found in sourceNo objects to start with in target Stop This will ensure that we include all the necessary objects to make any synchronization work. However, there is still the issue of query performance; the CONNECT BY on the entire database dependency graph is still too slow. After much sitting down and drawing complicated diagrams, we decided to move the graph traversal algorithm from the server onto the client (which turned out to run much faster on the client than on the server); and to ensure we don’t read the entire dependency graph onto the client we also pull the graph across in bits – we start off with dependency edges involving schemas selected for explicit population, and whenever the graph traversal comes across a dependency reference to a schema we don’t yet know about a thunk is hit that pulls in the dependency information for that schema from the database. We continue passing more dependent objects back and forth between the source and target until no more dependency references are found. This gives us the list of all the extra objects to populate in the source and target, and object population can then proceed. 4. Object blacklists and fast dependencies When we tested this solution, we were puzzled in that in some of our databases most of the system schemas (WMSYS, ORDSYS, EXFSYS, XDB, etc) were being pulled in, and this was increasing the database registration and comparison time quite significantly. After debugging, we discovered that the culprits were database tables that used one of the Oracle PL/SQL types (eg the SDO_GEOMETRY spatial type). These were creating a dependency chain from the database tables we were populating to the system schemas, and hence pulling in most of the system objects in that schema. To solve this we introduced blacklists of objects we wouldn’t follow any dependency chain through. As well as the Oracle-supplied PL/SQL types (MDSYS.SDO_GEOMETRY, ORDSYS.SI_COLOR, among others) we also decided to blacklist the entire PUBLIC and SYS schemas, as any references to those would likely lead to a blow up in the dependency graph that would massively increase the database registration time, and could result in the client running out of memory. Even with these improvements, each dependency query was taking upwards of a minute. We discovered from Oracle execution plans that there were some columns, with dependency information we required, that were querying system tables with no indexes on them! To cut a long story short, running the following query: SELECT * FROM all_tab_cols WHERE data_type_owner = ‘XDB’; results in a full table scan of the SYS.COL$ system table! This single clause was responsible for over half the execution time of the dependency query. Hence, the ‘Ignore slow dependencies’ option was born – not querying this and a couple of similar clauses to drastically speed up the dependency query execution time, at the expense of producing incorrect sync scripts in rare edge cases. Needless to say, along with the sync script action ordering, the dependency code in the database registration is one of the most complicated and most rewritten parts of the Schema Compare for Oracle engine. The beta of Schema Compare for Oracle is out now; if you find a bug in it, please do tell us so we can get it fixed!

    Read the article

  • Crime Scene Investigation: SQL Server

    - by Rodney Landrum
    “The packages are running slower in Prod than they are in Dev” My week began with this simple declaration from one of our lead BI developers, quickly followed by an emailed spreadsheet demonstrating that, over 5 executions, an extensive ETL process was running average 630 seconds faster on Dev than on Prod. The situation needed some scientific investigation to determine why the same code, the same data, the same schema would yield consistently slower results on a more powerful server. Prod had yet to be officially christened with a “Go Live” date so I had the time, and having recently been binge watching CSI: New York, I also had the inclination. An inspection of the two systems, Prod and Dev, revealed the first surprise: although Prod was indeed a “bigger” system, with double the amount of RAM of Dev, the latter actually had twice as many processor cores. On neither system did I see much sign of resources being heavily taxed, while the ETL process was running. Without any real supporting evidence, I jumped to a conclusion that my years of performance tuning should have helped me avoid, and that was that the hardware differences explained the better performance on Dev. We spent time setting up a Test system, similarly scoped to Prod except with 4 times the cores, and ported everything across. The results of our careful benchmarks left us truly bemused; the ETL process on the new server was slower than on both other systems. We burned more time tweaking server configurations, monitoring IO and network latency, several times believing we’d uncovered the smoking gun, until the results of subsequent test runs pitched us back into confusion. Finally, I decided, enough was enough. Hadn’t I learned very early in my DBA career that almost all bottlenecks were caused by code and database design, not hardware? It was time to get back to basics. With over 100 SSIS packages and hundreds of queries, each handling specific tasks such as file loads, bulk inserts, transforms, logging, and so on, the task seemed formidable. And yet, after barely an hour spent with Profiler, Extended Events, and wait statistics DMVs, I had a lead in the shape of a query that joined three tables, containing millions of rows, returned 3279 results, but performed 239K logical reads. As soon as I looked at the execution plans for the query in Dev and Test I saw the culprit, an implicit conversion warning on a join predicate field that was numeric in one table and a varchar(50) in another! I turned this information over to the BI developers who quickly resolved the data type mismatches and found and fixed “several” others as well. After the schema changes the same query with the same databases ran in under 1 second on all systems and reduced the logical reads down to fewer than 300. The analysis also revealed that on Dev, the ETL task was pulling data across a LAN, whereas Prod and Test were connected across slower WAN, in large part explaining why the same process ran slower on the latter two systems. Loading the data locally on Prod delivered a further 20% gain in performance. As we progress through our DBA careers we learn valuable lessons. Sometimes, with a project deadline looming and pressure mounting, we choose to forget them. I was close to giving into the temptation to throw more hardware at the problem. I’m pleased at least that I resisted, though I still kick myself for not looking at the code on day one. It can seem a daunting prospect to return to the fundamentals of the code so close to roll out, but with the right tools, and surprisingly little time, you can collect the evidence that reveals the true problem. It is a lesson I trust I will remember for my next 20 years as a DBA, if I’m ever again tempted to bypass the evidence.

    Read the article

  • A weekend with the Samsung Galaxy Tab

    - by Richard Mitchell
    This weekend I took one of the Samsung Galaxy Tabs we have lying around the office here home to see how I got on with it as I've been thinking of buying one. Initial impressions The look and feel of the Tab is quite nice. It's a lot smaller than an iPad but that is no bad thing as I imagine they are targeted at different markets. The Tab fits into my inside coat pocket nicely and doesn't feel like it's weighing me down too much. Connecting up the Tab to the network at work was fine, typing in...(read more)

    Read the article

< Previous Page | 27 28 29 30 31 32 33 34 35 36 37 38  | Next Page >