Search Results

Search found 5853 results on 235 pages for 'vivian short'.

Page 206/235 | < Previous Page | 202 203 204 205 206 207 208 209 210 211 212 213  | Next Page >

  • Performance triage

    - by Dave
    Folks often ask me how to approach a suspected performance issue. My personal strategy is informed by the fact that I work on concurrency issues. (When you have a hammer everything looks like a nail, but I'll try to keep this general). A good starting point is to ask yourself if the observed performance matches your expectations. Expectations might be derived from known system performance limits, prototypes, and other software or environments that are comparable to your particular system-under-test. Some simple comparisons and microbenchmarks can be useful at this stage. It's also useful to write some very simple programs to validate some of the reported or expected system limits. Can that disk controller really tolerate and sustain 500 reads per second? To reduce the number of confounding factors it's better to try to answer that question with a very simple targeted program. And finally, nothing beats having familiarity with the technologies that underlying your particular layer. On the topic of confounding factors, as our technology stacks become deeper and less transparent, we often find our own technology working against us in some unexpected way to choke performance rather than simply running into some fundamental system limit. A good example is the warm-up time needed by just-in-time compilers in Java Virtual Machines. I won't delve too far into that particular hole except to say that it's rare to find good benchmarks and methodology for java code. Another example is power management on x86. Power management is great, but it can take a while for the CPUs to throttle up from low(er) frequencies to full throttle. And while I love "turbo" mode, it makes benchmarking applications with multiple threads a chore as you have to remember to turn it off and then back on otherwise short single-threaded runs may look abnormally fast compared to runs with higher thread counts. In general for performance characterization I disable turbo mode and fix the power governor at "performance" state. Another source of complexity is the scheduler, which I've discussed in prior blog entries. Lets say I have a running application and I want to better understand its behavior and performance. We'll presume it's warmed up, is under load, and is an execution mode representative of what we think the norm would be. It should be in steady-state, if a steady-state mode even exists. On Solaris the very first thing I'll do is take a set of "pstack" samples. Pstack briefly stops the process and walks each of the stacks, reporting symbolic information (if available) for each frame. For Java, pstack has been augmented to understand java frames, and even report inlining. A few pstack samples can provide powerful insight into what's actually going on inside the program. You'll be able to see calling patterns, which threads are blocked on what system calls or synchronization constructs, memory allocation, etc. If your code is CPU-bound then you'll get a good sense where the cycles are being spent. (I should caution that normal C/C++ inlining can diffuse an otherwise "hot" method into other methods. This is a rare instance where pstack sampling might not immediately point to the key problem). At this point you'll need to reconcile what you're seeing with pstack and your mental model of what you think the program should be doing. They're often rather different. And generally if there's a key performance issue, you'll spot it with a moderate number of samples. I'll also use OS-level observability tools to lock for the existence of bottlenecks where threads contend for locks; other situations where threads are blocked; and the distribution of threads over the system. On Solaris some good tools are mpstat and too a lesser degree, vmstat. Try running "mpstat -a 5" in one window while the application program runs concurrently. One key measure is the voluntary context switch rate "vctx" or "csw" which reflects threads descheduling themselves. It's also good to look at the user; system; and idle CPU percentages. This can give a broad but useful understanding if your threads are mostly parked or mostly running. For instance if your program makes heavy use of malloc/free, then it might be the case you're contending on the central malloc lock in the default allocator. In that case you'd see malloc calling lock in the stack traces, observe a high csw/vctx rate as threads block for the malloc lock, and your "usr" time would be less than expected. Solaris dtrace is a wonderful and invaluable performance tool as well, but in a sense you have to frame and articulate a meaningful and specific question to get a useful answer, so I tend not to use it for first-order screening of problems. It's also most effective for OS and software-level performance issues as opposed to HW-level issues. For that reason I recommend mpstat & pstack as my the 1st step in performance triage. If some other OS-level issue is evident then it's good to switch to dtrace to drill more deeply into the problem. Only after I've ruled out OS-level issues do I switch to using hardware performance counters to look for architectural impediments.

    Read the article

  • A Knights Tale

    - by Phil Factor
    There are so many lessons to be learned from the story of Knight Capital losing nearly half a billion dollars as a result of a deployment gone wrong. The Knight Capital Group (KCG N) was an American global financial services firm engaging in market making, electronic execution, and institutional sales and trading. According to the recent order (File No.3.15570) against Knight Capital by U.S. Securities and Exchange Commission?, Knight had, for many years used some software which broke up incoming “parent” orders into smaller “child” orders that were then transmitted to various exchanges or trading venues for execution. A tracking ‘cumulative quantity’ function counted the number of ‘child’ orders and stopped the process once the total of child orders matched the ‘parent’ and so the parent order had been completed. Back in the mists of time, some code had been added to it  which was excuted if a particular flag was set. It was called ‘power peg’ and seems to have had a similar design and purpose, but, one guesses, would have shared the same tracking function. This code had been abandoned in 2003, but never deleted. In 2005, The tracking function was moved to an earlier point in the main process. It would seem from the account that, from that point, had that flag ever been set, the old ‘Power Peg’ would have been executed like Godzilla bursting from the ice, making child orders without limit without any tracking function. It wasn’t, presumably because the software that set the flag was removed. In 2012, nearly a decade after ‘Power Peg’ was abandoned, Knight prepared a new module to their software to cope with the imminent Retail Liquidity Program (RLP) for the New York Stock Exchange. By this time, the flag had remained unused and someone made the fateful decision to reuse it, and replace the old ‘power peg’ code with this new RLP code. Had the two actions been done together in a single automated deployment, and the new deployment tested, all would have been well. It wasn’t. To quote… “Beginning on July 27, 2012, Knight deployed the new RLP code in SMARS in stages by placing it on a limited number of servers in SMARS on successive days. During the deployment of the new code, however, one of Knight’s technicians did not copy the new code to one of the eight SMARS computer servers. Knight did not have a second technician review this deployment and no one at Knight realized that the Power Peg code had not been removed from the eighth server, nor the new RLP code added. Knight had no written procedures that required such a review.” (para 15) “On August 1, Knight received orders from broker-dealers whose customers were eligible to participate in the RLP. The seven servers that received the new code processed these orders correctly. However, orders sent with the repurposed flag to the eighth server triggered the defective Power Peg code still present on that server. As a result, this server began sending child orders to certain trading centers for execution. Because the cumulative quantity function had been moved, this server continuously sent child orders, in rapid sequence, for each incoming parent order without regard to the number of share executions Knight had already received from trading centers. Although one part of Knight’s order handling system recognized that the parent orders had been filled, this information was not communicated to SMARS.” (para 16) SMARS routed millions of orders into the market over a 45-minute period, and obtained over 4 million executions in 154 stocks for more than 397 million shares. By the time that Knight stopped sending the orders, Knight had assumed a net long position in 80 stocks of approximately $3.5 billion and a net short position in 74 stocks of approximately $3.15 billion. Knight’s shares dropped more than 20% after traders saw extreme volume spikes in a number of stocks, including preferred shares of Wells Fargo (JWF) and semiconductor company Spansion (CODE). Both stocks, which see roughly 100,000 trade per day, had changed hands more than 4 million times by late morning. Ultimately, Knight lost over $460 million from this wild 45 minutes of trading. Obviously, I’m interested in all this because, at one time, I used to write trading systems for the City of London. Obviously, the US SEC is in a far better position than any of us to work out the failings of Knight’s IT department, and the report makes for painful reading. I can’t help observing, though, that even with the breathtaking mistakes all along the way, that a robust automated deployment process that was ‘all-or-nothing’, and tested from soup to nuts would have prevented the disaster. The report reads like a Greek Tragedy. All the way along one wants to shout ‘No! not that way!’ and ‘Aargh! Don’t do it!’. As the tragedy unfolds, the audience weeps for the players, trapped by a cruel fate. All application development and deployment requires defense in depth. All IT goes wrong occasionally, but if there is a culture of defensive programming throughout, the consequences are usually containable. For financial systems, these defenses are required by statute, and ignored only by the foolish. Knight’s mistakes weren’t made by just one hapless sysadmin, but were progressive errors by an  IT culture spanning at least ten years.  One can spell these out, but I think they’re obvious. One can only hope that the industry studies what happened in detail, learns from the mistakes, and draws the right conclusions.

    Read the article

  • What is Happening vs. What is Interesting

    - by Geertjan
    Devoxx 2011 was yet another confirmation that all development everywhere is either on the web or on mobile phones. Whether you looked at the conference schedule or attended sessions or talked to speakers at any point at all, it was very clear that no development whatsoever is done anymore on the desktop. In fact, that's something Tim Bray himself told me to my face at the speakers dinner. No new developments of any kind are happening on the desktop. Everyone who is currently on the desktop is working overtime to move all of their applications to the web. They're probably also creating a small subset of their application on an Android tablet, with an even smaller subset on their Android phone. Then you scratch that monolithic surface and find some interesting results. Without naming any names, I asked one of these prominent "ah, forget about the desktop" people at the Devoxx speakers dinner (and I have a witness): "Yes, the desktop is dead, but what about air traffic control, stock trading, oil analysis, risk management applications? In fact, what about any back office application that needs to be usable across all operating systems? Here there is no concern whatsoever with 100% accessibility which is, after all, the only thing that the web has over the desktop, (except when there's a network failure, of course, or when you find yourself in the 3/4 of the world where there's bandwidth problems)? There are 1000's of hidden applications out there that have processing requirements, security requirements, and the requirement that they'll be available even when the network is down or even completely unavailable. Isn't that a valid use case and aren't there 1000's of applications that fall into this so-called niche category? Are you not, in fact, confusing consumer applications, which are increasingly web-based and mobile-based, with high-end corporate applications, which typically need to do massive processing, of one kind or another, for which the web and mobile worlds are completely unsuited?" And you will not believe what the reply to the above question was. (Again, I have a witness to this discussion.) But here it is: "Yes. But those applications are not interesting. I do not want to spend any of my time or work in any way on those applications. They are boring." I'm sad to say that the leaders of the software development community, including those in the Java world, either share the above opinion or are led by it. Because they find something that is not new to be boring, they move on to what is interesting and start talking like the supposedly-boring developments don't even exist. (Kind of like a rapper pretending classical music doesn't exist.) Time and time again I find myself giving Java desktop development courses (at companies, i.e., not hobbyists, or students, but companies, i.e., the places where dollars are earned), where developers say to me: "The course you're giving about creating cross-platform, loosely coupled, and highly cohesive applications is really useful to us. Why do we never find information about this topic at conferences? Why can we never attend a session at a conference where the story about pluggable cross-platform Java is told? Why do we get the impression that we are uncool because we're not on the web and because we're not on a mobile phone, while the reason for that is because we're creating $1000,000 simulation software which has nothing to gain from being on the web or on the mobile phone?" And then I say: "Because nobody knows you exist. Because you're not submitting abstracts to conferences about your very interesting use cases. And because conferences tend to focus on what is new, which tends to be web related (especially HTML 5) or mobile related (especially Android). Because you're not taking the responsibility on yourself to tell the real stories about the real applications being developed all the time and every day. Because you yourself think your work is boring, while in fact it is fascinating. Because desktop developers are working from 9 to 5 on the desktop, in secure environments, such as banks and defense, where you can't spend time, nor have the interest in, blogging your latest tip or trick, as opposed to web developers, who tend to spend a lot of time on the web anyway and are therefore much more inclined to create buzz about the kind of work they're doing." So, next time you look at a conference program and wonder why there's no stories about large desktop development projects in the program, here's the short answer: "No one is going to put those items on the program until you start submitting those kinds of sessions. And until you start blogging. Until you start creating the buzz that the web developers have been creating around their work for the past 10 years or so. And, yes, indeed, programmers get the conference they deserve." And what about Tim Bray? Ask yourself, as Google's lead web technology evangelist, how many desktop developers do you think he talks to and, more generally, what his frame of reference is and what, clearly, he considers to be most interesting.

    Read the article

  • ADF Business Components

    - by Arda Eralp
    ADF Business Components and JDeveloper simplify the development, delivery, and customization of business applications for the Java EE platform. With ADF Business Components, developers aren't required to write the application infrastructure code required by the typical Java EE application to: Connect to the database Retrieve data Lock database records Manage transactions   ADF Business Components addresses these tasks through its library of reusable software components and through the supporting design time facilities in JDeveloper. Most importantly, developers save time using ADF Business Components since the JDeveloper design time makes typical development tasks entirely declarative. In particular, JDeveloper supports declarative development with ADF Business Components to: Author and test business logic in components which automatically integrate with databases Reuse business logic through multiple SQL-based views of data, supporting different application tasks Access and update the views from browser, desktop, mobile, and web service clients Customize application functionality in layers without requiring modification of the delivered application The goal of ADF Business Components is to make the business services developer more productive.   ADF Business Components provides a foundation of Java classes that allow your business-tier application components to leverage the functionality provided in the following areas: Simplifying Data Access Design a data model for client displays, including only necessary data Include master-detail hierarchies of any complexity as part of the data model Implement end-user Query-by-Example data filtering without code Automatically coordinate data model changes with business services layer Automatically validate and save any changes to the database   Enforcing Business Domain Validation and Business Logic Declaratively enforce required fields, primary key uniqueness, data precision-scale, and foreign key references Easily capture and enforce both simple and complex business rules, programmatically or declaratively, with multilevel validation support Navigate relationships between business domain objects and enforce constraints related to compound components   Supporting Sophisticated UIs with Multipage Units of Work Automatically reflect changes made by business service application logic in the user interface Retrieve reference information from related tables, and automatically maintain the information when the user changes foreign-key values Simplify multistep web-based business transactions with automatic web-tier state management Handle images, video, sound, and documents without having to use code Synchronize pending data changes across multiple views of data Consistently apply prompts, tooltips, format masks, and error messages in any application Define custom metadata for any business components to support metadata-driven user interface or application functionality Add dynamic attributes at runtime to simplify per-row state management   Implementing High-Performance Service-Oriented Architecture Support highly functional web service interfaces for business integration without writing code Enforce best-practice interface-based programming style Simplify application security with automatic JAAS integration and audit maintenance "Write once, run anywhere": use the same business service as plain Java class, EJB session bean, or web service   Streamlining Application Customization Extend component functionality after delivery without modifying source code Globally substitute delivered components with extended ones without modifying the application   ADF Business Components implements the business service through the following set of cooperating components: Entity object An entity object represents a row in a database table and simplifies modifying its data by handling all data manipulation language (DML) operations for you. These are basically your 1 to 1 representation of a database table. Each table in the database will have 1 and only 1 EO. The EO contains the mapping between columns and attributes. EO's also contain the business logic and validation. These are you core data services. They are responsible for updating, inserting and deleting records. The Attributes tab displays the actual mapping between attributes and columns, the mapping has following fields: Name : contains the name of the attribute we expose in our data model. Type : defines the data type of the attribute in our application. Column : specifies the column to which we want to map the attribute with Column Type : contains the type of the column in the database   View object A view object represents a SQL query. You use the full power of the familiar SQL language to join, filter, sort, and aggregate data into exactly the shape required by the end-user task. The attributes in the View Objects are actually coming from the Entity Object. In the end the VO will generate a query but you basically build a VO by selecting which EO need to participate in the VO and which attributes of those EO you want to use. That's why you have the Entity Usage column so you can see the relation between VO and EO. In the query tab you can clearly see the query that will be generated for the VO. At this stage we don't need it and just use it for information purpose. In later stages we might use it. Application module An application module is the controller of your data layer. It is responsible for keeping hold of the transaction. It exposes the data model to the view layer. You expose the VO's through the Application Module. This is the abstraction of your data layer which you want to show to the outside word.It defines an updatable data model and top-level procedures and functions (called service methods) related to a logical unit of work related to an end-user task. While the base components handle all the common cases through built-in behavior, customization is always possible and the default behavior provided by the base components can be easily overridden or augmented. When you create EO's, a foreign key will be translated into an association in our model. It defines the type of relation and who is the master and child as well as how the visibility of the association looks like. A similar concept exists to identify relations between view objects. These are called view links. These are almost identical as association except that a view link is based upon attributes defined in the view object. It can also be based upon an association. Here's a short summary: Entity Objects: representations of tables Association: Relations between EO's. Representations of foreign keys View Objects: Logical model View Links: Relationships between view objects Application Model: interface to your application  

    Read the article

  • Separating text strings into a table of individual words in SQL via XML.

    - by Phil Factor
    p.MsoNormal {margin-top:0cm; margin-right:0cm; margin-bottom:10.0pt; margin-left:0cm; line-height:115%; font-size:11.0pt; font-family:"Calibri","sans-serif"; } Nearly nine years ago, Mike Rorke of the SQL Server 2005 XML team blogged ‘Querying Over Constructed XML Using Sub-queries’. I remember reading it at the time without being able to think of a use for what he was demonstrating. Just a few weeks ago, whilst preparing my article on searching strings, I got out my trusty function for splitting strings into words and something reminded me of the old blog. I’d been trying to think of a way of using XML to split strings reliably into words. The routine I devised turned out to be slightly slower than the iterative word chop I’ve always used in the past, so I didn’t publish it. It was then I suddenly remembered the old routine. Here is my version of it. I’ve unwrapped it from its obvious home in a function or procedure just so it is easy to appreciate. What it does is to chop a text string into individual words using XQuery and the good old nodes() method. I’ve benchmarked it and it is quicker than any of the SQL ways of doing it that I know about. Obviously, you can’t use the trick I described here to do it, because it is awkward to use REPLACE() on 1…n characters of whitespace. I’ll carry on using my iterative function since it is able to tell me the location of each word as a character-offset from the start, and also because this method leaves punctuation in (removing it takes time!). However, I can see other uses for this in passing lists as input or output parameters, or as return values.   if exists (Select * from sys.xml_schema_collections where name like 'WordList')   drop XML SCHEMA COLLECTION WordList go create xml schema collection WordList as ' <xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"> <xs:element name="words">        <xs:simpleType>               <xs:list itemType="xs:string" />        </xs:simpleType> </xs:element> </xs:schema>'   go   DECLARE @string VARCHAR(MAX) –we'll get some sample data from the great Ogden Nash Select @String='This is a song to celebrate banks, Because they are full of money and you go into them and all you hear is clinks and clanks, Or maybe a sound like the wind in the trees on the hills, Which is the rustling of the thousand dollar bills. Most bankers dwell in marble halls, Which they get to dwell in because they encourage deposits and discourage withdrawals, And particularly because they all observe one rule which woe betides the banker who fails to heed it, Which is you must never lend any money to anybody unless they don''t need it. I know you, you cautious conservative banks! If people are worried about their rent it is your duty to deny them the loan of one nickel, yes, even one copper engraving of the martyred son of the late Nancy Hanks; Yes, if they request fifty dollars to pay for a baby you must look at them like Tarzan looking at an uppity ape in the jungle, And tell them what do they think a bank is, anyhow, they had better go get the money from their wife''s aunt or ungle. But suppose people come in and they have a million and they want another million to pile on top of it, Why, you brim with the milk of human kindness and you urge them to accept every drop of it, And you lend them the million so then they have two million and this gives them the idea that they would be better off with four, So they already have two million as security so you have no hesitation in lending them two more, And all the vice-presidents nod their heads in rhythm, And the only question asked is do the borrowers want the money sent or do they want to take it withm. Because I think they deserve our appreciation and thanks, the jackasses who go around saying that health and happi- ness are everything and money isn''t essential, Because as soon as they have to borrow some unimportant money to maintain their health and happiness they starve to death so they can''t go around any more sneering at good old money, which is nothing short of providential. '   –we now turn it into XML declare @xml_data xml(WordList)  set @xml_data='<words>'+ replace(@string,'&', '&amp;')+'</words>'    select T.ref.value('.', 'nvarchar(100)')  from (Select @xml_data.query('                      for $i in data(/words) return                      element li { $i }               '))  A(list) cross apply A.List.nodes('/li') T(ref)     …which gives (truncated, of course)…

    Read the article

  • iOS Support with Windows Azure Mobile Services – now with Push Notifications

    - by ScottGu
    A few weeks ago I posted about a number of improvements to Windows Azure Mobile Services. One of these was the addition of an Objective-C client SDK that allows iOS developers to easily use Mobile Services for data and authentication.  Today I'm excited to announce a number of improvement to our iOS SDK and, most significantly, our new support for Push Notifications via APNS (Apple Push Notification Services).  This makes it incredibly easy to fire push notifications to your iOS users from Windows Azure Mobile Service scripts. Push Notifications via APNS We've provided two complete tutorials that take you step-by-step through the provisioning and setup process to enable your Windows Azure Mobile Service application with APNS (Apple Push Notification Services), including all of the steps required to configure your application for push in the Apple iOS provisioning portal: Getting started with Push Notifications - iOS Push notifications to users by using Mobile Services - iOS Once you've configured your application in the Apple iOS provisioning portal and uploaded the APNS push certificate to the Apple provisioning portal, it's just a matter of uploading your APNS push certificate to Mobile Services using the Windows Azure admin portal: Clicking the “upload” within the “Push” tab of your Mobile Service allows you to browse your local file-system and locate/upload your exported certificate.  As part of this you can also select whether you want to use the sandbox (dev) or production (prod) Apple service: Now, the code to send a push notification to your clients from within a Windows Azure Mobile Service is as easy as the code below: push.apns.send(deviceToken, {      alert: 'Toast: A new Mobile Services task.',      sound: 'default' }); This will cause Windows Azure Mobile Services to connect to APNS (Apple Push Notification Service) and send a notification to the iOS device you specified via the deviceToken: Check out our reference documentation for full details on how to use the new Windows Azure Mobile Services apns object to send your push notifications. Feedback Scripts An important part of working with any PNS (Push Notification Service) is handling feedback for expired device tokens and channels. This typically happens when your application is uninstalled from a particular device and can no longer receive your notifications. With Windows Notification Services you get an instant response from the HTTP server.  Apple’s Notification Services works in a slightly different way and provides an additional endpoint you can connect to poll for a list of expired tokens. As with all of the capabilities we integrate with Mobile Services, our goal is to allow developers to focus more on building their app and less on building infrastructure to support their ideas. Therefore we knew we had to provide a simple way for developers to integrate feedback from APNS on a regular basis.  This week’s update now includes a new screen in the portal that allows you to optionally provide a script to process your APNS feedback – and it will be executed by Mobile Services on an ongoing basis: This script is invoked periodically while your service is active. To poll the feedback endpoint you can simply call the apns object's getFeedback method from within this script: push.apns.getFeedback({       success: function(results) {           // results is an array of objects with a deviceToken and time properties      } }); This returns you a list of invalid tokens that can now be removed from your database. iOS Client SDK improvements Over the last month we've continued to work with a number of iOS advisors to make improvements to our Objective-C SDK. The SDK is being developed under an open source license (Apache 2.0) and is available on github. Many of the improvements are behind the scenes to improve performance and memory usage. However, one of the biggest improvements to our iOS Client API is the addition of an even easier login method.  Below is the Objective-C code you can now write to invoke it: [client loginWithProvider:@"twitter"                     onController:self                        animated:YES                      completion:^(MSUser *user, NSError *error) {      // if no error, you are now logged in via twitter }]; This code will automatically present and dismiss our login view controller as a modal dialog on the specified controller.  This does all the hard work for you and makes login via Twitter, Google, Facebook and Microsoft Account identities just a single line of code. My colleague Josh just posted a short video demonstrating these new features which I'd recommend checking out: Summary The above features are all now live in production and are available to use immediately.  If you don’t already have a Windows Azure account, you can sign-up for a free trial and start using Mobile Services today. Visit the Windows Azure Mobile Developer Center to learn more about how to build apps with Mobile Services. Hope this helps, Scott P.S. In addition to blogging, I am also now using Twitter for quick updates and to share links. Follow me at: twitter.com/scottgu

    Read the article

  • Microsoft Sql Server 2008 R2 System Databases

    For a majority of software developers little time is spent understanding the inner workings of the database management systems (DBMS) they use to store data for their applications.  I personally place myself in this grouping. In my case, I have used various versions of Microsoft’s SQL Server (2000, 2005, and 2008 R2) and just recently learned how valuable they really are when I was preparing to deliver a lecture on "SQL Server 2008 R2, System Databases". Microsoft Sql Server 2008 R2 System DatabasesSo what are system databases in MS SQL Server, and why should I know them? Microsoft uses system databases to support the SQL Server DBMS, much like a developer uses config files or database tables to support an application. These system databases individually provide specific functionality that allows MS SQL Server to function. Name Database File Log File Master master.mdf mastlog.ldf Resource mssqlsystemresource.mdf mssqlsystemresource.ldf Model model.mdf modellog.ldf MSDB msdbdata.mdf msdblog.ldf Distribution distmdl.mdf distmdl.ldf TempDB tempdb.mdf templog.ldf Master DatabaseIf you have used MS SQL Server then you should recognize the Master database especially if you used the SQL Server Management Studio (SSMS) to connect to a user created database. MS SQL Server requires the Master database in order for DBMS to start due to the information that it stores. Examples of data stored in the Master database User Logins Linked Servers Configuration information Information on User Databases Resource DatabaseHonestly, until recently I never knew this database even existed until I started to research SQL Server system databases. The reason for this is due largely to the fact that the resource database is hidden to users. In fact, the database files are stored within the Binn folder instead of the standard MS SQL Server database folder path. This database contains all system objects that can be accessed by all other databases.  In short, this database contains all system views and store procedures that appear in all other user databases regarding system information. One of the many benefits to storing system views and store procedures in a single hidden database is the fact it improves upgrading a SQL Server database; not to mention that maintenance is decreased since only one code base has to be mainlined for all of the system views and procedures. Model DatabaseThe Model database as the name implies is the model for all new databases created by users. This allows for predefining default database objects for all new databases within a MS SQL Server instance. For example, if every database created by a user needs to have an “Audit” table when it is  created then defining the “Audit” table in the model will guarantees that the table will be located in every new database create after the model is altered. MSDB DatabaseThe MSDBdatabase is used by SQL Server Agent, SQL Server Database Mail, SQL Server Service Broker, along with SQL Server. The SQL Server Agent uses this database to store job configurations and SQL job schedules along with SQL Alerts, and Operators. In addition, this database also stores all SQL job parameters along with each job’s execution history.  Finally, this database is also used to store database backup and maintenance plans as well as details pertaining to SQL Log shipping if it is being used. Distribution DatabaseThe Distribution database is only used during replication and stores meta data and history information pertaining to the act of replication data. Furthermore, when transactional replication is used this database also stores information regarding each transaction. It is important to note that replication is not turned on by default in MS SQL Server and that the distribution database is hidden from SSMS. Tempdb DatabaseThe Tempdb as the name implies is used to store temporary data and data objects. Examples of this include temp tables and temp store procedures. It is important to note that when using this database all data and data objects are cleared from this database when SQL Server restarts. This database is also used by SQL Server when it is performing some internal operations. Typically, SQL Server uses this database for the purpose of large sort and index operations. Finally, this database is used to store row versions if row versioning or snapsot isolation transactions are being used by SQL Server. Additionally, I would love to hear from others about their experiences using system databases, tables, and objects in a real world environments.

    Read the article

  • Why won't USB 3.0 external hard drive run at USB 3.0 speeds?

    - by jgottula
    I recently purchased a PCI Express x1 USB 3.0 controller card (containing the NEC USB 3.0 controller) with the intent of using a USB 3.0 external hard drive with my Linux box. I installed the card in an empty PCIe slot on my motherboard, connected the card to a power cable, strung a USB 3.0 cable between one of the new ports and my external HDD, and connected the HDD to a wall socket for power. Booting the system, the drive works 100% as intended, with the one exception of throughput: rather than using SuperSpeed 4.8 Gbps connectivity, it seems to be falling back to High Speed 480 Mbps USB 2.0-style throughput. Disk Utility shows it as a 480 Mbps device, and running a couple Disk Utility and dd benchmarks confirms that the drive fails to exceed ~40 MB/s (the approximate limit of USB 2.0), despite it being an SSD capable of far more than that. When I connect my USB 3.0 HDD, dmesg shows this: [ 3923.280018] usb 3-2: new high speed USB device using ehci_hcd and address 6 where I would expect to find this: [ 3923.280018] usb 3-2: new SuperSpeed USB device using xhci_hcd and address 6 My system was running on kernel 2.6.35-25-generic at the time. Then, I stumbled upon this forum thread by an individual who found that a bug, which was present in kernels prior to 2.6.37-rc5, could be the culprit for this type of problem. Consequently, I installed the 2.6.37-generic mainline Ubuntu kernel to determine if the problem would go away. It didn't, so I tried 2.6.38-rc3-generic, and even the 2.6.38 nightly from 2010.02.01, to no avail. In short, I'm trying to determine why, with USB 3.0 support in the kernel, my USB 3.0 drive fails to run at full SuperSpeed throughput. See the comments under this question for additional details. Output that might be relevant to the problem (when booting from 2.6.38-rc3): Relevant lines from dmesg: [ 19.589491] xhci_hcd 0000:03:00.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17 [ 19.589512] xhci_hcd 0000:03:00.0: setting latency timer to 64 [ 19.589516] xhci_hcd 0000:03:00.0: xHCI Host Controller [ 19.589623] xhci_hcd 0000:03:00.0: new USB bus registered, assigned bus number 12 [ 19.650492] xhci_hcd 0000:03:00.0: irq 17, io mem 0xf8100000 [ 19.650556] xhci_hcd 0000:03:00.0: irq 47 for MSI/MSI-X [ 19.650560] xhci_hcd 0000:03:00.0: irq 48 for MSI/MSI-X [ 19.650563] xhci_hcd 0000:03:00.0: irq 49 for MSI/MSI-X [ 19.653946] xHCI xhci_add_endpoint called for root hub [ 19.653948] xHCI xhci_check_bandwidth called for root hub Relevant section of sudo lspci -v: 03:00.0 USB Controller: NEC Corporation uPD720200 USB 3.0 Host Controller (rev 03) (prog-if 30) Flags: bus master, fast devsel, latency 0, IRQ 17 Memory at f8100000 (64-bit, non-prefetchable) [size=8K] Capabilities: [50] Power Management version 3 Capabilities: [70] MSI: Enable- Count=1/8 Maskable- 64bit+ Capabilities: [90] MSI-X: Enable+ Count=8 Masked- Capabilities: [a0] Express Endpoint, MSI 00 Capabilities: [100] Advanced Error Reporting Capabilities: [140] Device Serial Number ff-ff-ff-ff-ff-ff-ff-ff Capabilities: [150] #18 Kernel driver in use: xhci_hcd Kernel modules: xhci-hcd Relevant section of sudo lsusb -v: Bus 012 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub Device Descriptor: bLength 18 bDescriptorType 1 bcdUSB 3.00 bDeviceClass 9 Hub bDeviceSubClass 0 Unused bDeviceProtocol 3 bMaxPacketSize0 9 idVendor 0x1d6b Linux Foundation idProduct 0x0003 3.0 root hub bcdDevice 2.06 iManufacturer 3 Linux 2.6.38-020638rc3-generic xhci_hcd iProduct 2 xHCI Host Controller iSerial 1 0000:03:00.0 bNumConfigurations 1 Configuration Descriptor: bLength 9 bDescriptorType 2 wTotalLength 25 bNumInterfaces 1 bConfigurationValue 1 iConfiguration 0 bmAttributes 0xe0 Self Powered Remote Wakeup MaxPower 0mA Interface Descriptor: bLength 9 bDescriptorType 4 bInterfaceNumber 0 bAlternateSetting 0 bNumEndpoints 1 bInterfaceClass 9 Hub bInterfaceSubClass 0 Unused bInterfaceProtocol 0 Full speed (or root) hub iInterface 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x81 EP 1 IN bmAttributes 3 Transfer Type Interrupt Synch Type None Usage Type Data wMaxPacketSize 0x0004 1x 4 bytes bInterval 12 Hub Descriptor: bLength 9 bDescriptorType 41 nNbrPorts 4 wHubCharacteristic 0x0009 Per-port power switching Per-port overcurrent protection TT think time 8 FS bits bPwrOn2PwrGood 10 * 2 milli seconds bHubContrCurrent 0 milli Ampere DeviceRemovable 0x00 PortPwrCtrlMask 0xff Hub Port Status: Port 1: 0000.0100 power Port 2: 0000.0100 power Port 3: 0000.0100 power Port 4: 0000.0100 power Device Status: 0x0003 Self Powered Remote Wakeup Enabled Full, non-verbose lsusb: Bus 012 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub Bus 011 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 010 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 009 Device 003: ID 04d9:0702 Holtek Semiconductor, Inc. Bus 009 Device 002: ID 046d:c068 Logitech, Inc. G500 Laser Mouse Bus 009 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 008 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 006 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub Bus 003 Device 006: ID 174c:5106 ASMedia Technology Inc. Bus 003 Device 004: ID 0bda:0151 Realtek Semiconductor Corp. Mass Storage Device (Multicard Reader) Bus 003 Device 002: ID 058f:6366 Alcor Micro Corp. Multi Flash Reader Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 002 Device 006: ID 1687:0163 Kingmax Digital Inc. Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Bus 001 Device 002: ID 046d:081b Logitech, Inc. Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub Full output: full dmesg full lspci full lsusb

    Read the article

  • Clouds, Clouds, Clouds Everywhere, Not a Drop of Rain!

    - by sxkumar
    At the recently concluded Oracle OpenWorld 2012, the center of discussion was clearly Cloud. Over the five action packed days, I got to meet a large number of customers and most of them had serious interest in all things cloud.  Public Cloud - particularly the Oracle Cloud - clearly got a lot of attention and interest. I think the use cases and the value proposition for public cloud is pretty straight forward. However, when it comes to private cloud, there were some interesting revelations.  Well, I shouldn’t really call them revelations since they are pretty consistent with what I have heard from customers at other conferences as well as during 1:1 interactions. While the interest in enterprise private cloud remains to be very high, only a handful of enterprises have truly embarked on a journey to create what the purists would call true private cloud - with capabilities such as self-service and chargeback/show back. For a large majority, today's reality is simply consolidation and virtualization - and they are quite far off from creating an agile, self-service and transparent IT infrastructure which is what the enterprise cloud is all about.  Even a handful of those who have actually implemented a close-to-real enterprise private cloud have taken an infrastructure centric approach and are seeing only limited business upside. Quite a few were frank enough to admit that chargeback and self-service isn’t something that they see an immediate need for.  This is in quite contrast to the picture being painted by all those surveys out there that show a large number of enterprises having already implemented an enterprise private cloud.  On the face of it, this seems quite contrary to the observations outlined above. So what exactly is the reality? Well, the reality is that there is undoubtedly a huge amount of interest among enterprises about transforming their legacy IT environment - which is often seen as too rigid, too fragmented, and ultimately too expensive - to something more agile, transparent and business-focused. At the same time however, there is a great deal of confusion among CIOs and architects about how to get there. This isn't very surprising given all the buzz and hype surrounding cloud computing. Every IT vendor claims to have the most unique solution and there isn't a single IT product out there that does not have a cloud angle to it. Add to this the chatter on the blogosphere, it will get even a sane mind spinning.  Consequently, most  enterprises are still struggling to fully understand the concept and value of enterprise private cloud.  Even among those who have chosen to move forward relatively early, quite a few have made their decisions more based on vendor influence/preferences rather than what their businesses actually need.  Clearly, there is a disconnect between the promise of the enterprise private cloud and the current adoption trends.  So what is the way forward?  I certainly do not claim to have all the answers. But here is a perspective that many cloud practitioners have found useful and thus worth sharing. To take a step back, the fundamental premise of the enterprise private cloud is IT transformation. It is the quest to create a more agile, transparent and efficient IT infrastructure that is driven more by business needs rather than constrained by operational and procedural inefficiencies. It is the new way of delivering and consuming IT services - where the IT organizations operate more like enablers of  strategic services rather than just being the gatekeepers of IT resources. In an enterprise private cloud environment, IT organizations are expected to empower the end users via self-service access/control and provide the business stakeholders a transparent view of how the resources are being used, what’s the cost of delivering a given service, how well are the customers being served, etc.  But the most important thing to note here is the enterprise private cloud is not just an IT project, rather it is a business initiative to create an IT setup that is more aligned with the needs of today's dynamic and highly competitive business environment. Surprised? You shouldn’t be. Just remember how the business users have been at the forefront of public cloud adoption within enterprises and private cloud is no exception.   Such a broad-based transformation makes cloud more than a technology initiative. It requires people (organizational) and process changes as well, and these changes are as critical as is the choice of right tools and technology. In my next blog,  I will share how essential it is for enterprise cloud technology to go hand-in hand with process re-engineering and organization changes to unlock true value of  enterprise cloud. I am sharing a short video from my session "Managing your private Cloud" at Oracle OpenWorld 2012. More videos from this session will be posted at the recently introduced Zero to Cloud resource page. Many other experts of Oracle enterprise private cloud solution will join me on this blog "Zero to Cloud"  and share best practices , deployment tips and information on how to plan, build, deploy, monitor, manage , meter and optimize the enterprise private cloud. We look forward to your feedback, suggestions and having an engaging conversion with you on this blog.

    Read the article

  • Microsoft Sql Server 2008 R2 System Databases

    For a majority of software developers little time is spent understanding the inner workings of the database management systems (DBMS) they use to store data for their applications.  I personally place myself in this grouping. In my case, I have used various versions of Microsoft’s SQL Server (2000, 2005, and 2008 R2) and just recently learned how valuable they really are when I was preparing to deliver a lecture on "SQL Server 2008 R2, System Databases". Microsoft Sql Server 2008 R2 System DatabasesSo what are system databases in MS SQL Server, and why should I know them? Microsoft uses system databases to support the SQL Server DBMS, much like a developer uses config files or database tables to support an application. These system databases individually provide specific functionality that allows MS SQL Server to function. Name Database File Log File Master master.mdf mastlog.ldf Resource mssqlsystemresource.mdf mssqlsystemresource.ldf Model model.mdf modellog.ldf MSDB msdbdata.mdf msdblog.ldf Distribution distmdl.mdf distmdl.ldf TempDB tempdb.mdf templog.ldf Master DatabaseIf you have used MS SQL Server then you should recognize the Master database especially if you used the SQL Server Management Studio (SSMS) to connect to a user created database. MS SQL Server requires the Master database in order for DBMS to start due to the information that it stores. Examples of data stored in the Master database User Logins Linked Servers Configuration information Information on User Databases Resource DatabaseHonestly, until recently I never knew this database even existed until I started to research SQL Server system databases. The reason for this is due largely to the fact that the resource database is hidden to users. In fact, the database files are stored within the Binn folder instead of the standard MS SQL Server database folder path. This database contains all system objects that can be accessed by all other databases.  In short, this database contains all system views and store procedures that appear in all other user databases regarding system information. One of the many benefits to storing system views and store procedures in a single hidden database is the fact it improves upgrading a SQL Server database; not to mention that maintenance is decreased since only one code base has to be mainlined for all of the system views and procedures. Model DatabaseThe Model database as the name implies is the model for all new databases created by users. This allows for predefining default database objects for all new databases within a MS SQL Server instance. For example, if every database created by a user needs to have an “Audit” table when it is  created then defining the “Audit” table in the model will guarantees that the table will be located in every new database create after the model is altered. MSDB DatabaseThe MSDBdatabase is used by SQL Server Agent, SQL Server Database Mail, SQL Server Service Broker, along with SQL Server. The SQL Server Agent uses this database to store job configurations and SQL job schedules along with SQL Alerts, and Operators. In addition, this database also stores all SQL job parameters along with each job’s execution history.  Finally, this database is also used to store database backup and maintenance plans as well as details pertaining to SQL Log shipping if it is being used. Distribution DatabaseThe Distribution database is only used during replication and stores meta data and history information pertaining to the act of replication data. Furthermore, when transactional replication is used this database also stores information regarding each transaction. It is important to note that replication is not turned on by default in MS SQL Server and that the distribution database is hidden from SSMS. Tempdb DatabaseThe Tempdb as the name implies is used to store temporary data and data objects. Examples of this include temp tables and temp store procedures. It is important to note that when using this database all data and data objects are cleared from this database when SQL Server restarts. This database is also used by SQL Server when it is performing some internal operations. Typically, SQL Server uses this database for the purpose of large sort and index operations. Finally, this database is used to store row versions if row versioning or snapsot isolation transactions are being used by SQL Server. Additionally, I would love to hear from others about their experiences using system databases, tables, and objects in a real world environments.

    Read the article

  • The sign of a true manager is delegation (C# style)

    - by MarkPearl
    Today I thought I would write a bit about delegates in C#. Up till recently I have managed to side step any real understanding of what delegates do and why they are useful – I mean, I know roughly what they do and have used them a lot, but I have never really got down dirty with them and mucked about. Recently however with my renewed interest in Silverlight delegates came up again as a possible solution to a particular problem, and suddenly I found myself opening a bland little console application to just see exactly how far I could take delegates with my limited knowledge. So, let’s first look at the MSDN definition of delegates… A delegate declaration defines a reference type that can be used to encapsulate a method with a specific signature. A delegate instance encapsulates a static or an instance method. Delegates are roughly similar to function pointers in C++; however, delegates are type-safe and secure. Well, don’t you love MSDN for such a useful definition. I must give it credit though… later on it really explains it a bit better by saying “A delegate lets you pass a function as a parameter. The type safety of delegates requires the function you pass as a delegate to have the same signature as the delegate declaration.” A little more reading up on delegates mentions that delegates are similar to interfaces in that they enable the separation of specification and implementation. A delegate declares a single method, while an interface declares a group of methods. So enough reading - lets look at some code and see a basic example of a delegate… Let’s assume we have a console application with a simple delegate declared called AdjustValue like below… class Program { private delegate int AdjustValue(int val); static void Main(string[] args) { } } In a sense, all we have said is that we will be creating one or more methods that follow the same pattern as AdjustValue – i.e. they will take one input value of type int and return an integer. We could then expand our code to have various methods that match the structure of our delegate AdjustValue (remember the structure is int xxx (int xxx)) class Program { private delegate int AdjustValue(int val); private static int Dbl(int val) { return val * 2; } private static int AlwaysOne(int val) { return 1; } static void Main(string[] args) { } }  Above I have expanded my project to have two methods, one called Dbl and the other AlwaysOne. Dbl always returns double the input val and AlwaysOne always returns 1. I could now declare a variable and assign it to be one of those functions, like the following… class Program { private delegate int AdjustValue(int val); private static int Dbl(int val) { return val * 2; } private static int AlwaysOne(int val) { return 1; } static void Main(string[] args) { AdjustValue myDelegate; myDelegate = Dbl; Console.WriteLine(myDelegate(1).ToString()); Console.ReadLine(); } } In this instance I have declared an instance of the AdjustValue delegate called myDelegate; I have then told myDelegate to point to the method Dbl, and then called myDelegate(1). What would the result be? Yes, in this instance it would be exactly the same as me calling the following code… static void Main(string[] args) { Console.WriteLine(Dbl(1).ToString()); Console.ReadLine(); }   So why all the extra work for delegates when we could just do what we did above and call the method directly? Well… that separation of specification to implementation comes to mind. So, this all seems pretty simple. Let’s take a slightly more complicated variation to the console application. Assume that my project is the same as the one previously except that my main method is adjusted as follows… static void Main(string[] args) { AdjustValue myDelegate; myDelegate = Dbl; myDelegate = AlwaysOne; Console.WriteLine(myDelegate(1).ToString()); Console.ReadLine(); } What would happen in this scenario? Quite simply “1” would be written to the console, the reason being that myDelegate was last pointing to the AlwaysOne method before it was called. Make sense? In a way, the myDelegate is a variable method that can be swapped and changed when needed. Let’s make the code a little more confusing by using a delegate in the declaration of another delegate as shown below… class Program { private delegate int AdjustValue(InputValue val); private delegate int InputValue(); private static int Dbl(InputValue val) { return val()*2; } private static int GetInputVal() { Console.WriteLine("Enter a whole number : "); return Convert.ToInt32(Console.ReadLine()); } static void Main(string[] args) { AdjustValue myDelegate; myDelegate = Dbl; Console.WriteLine(myDelegate(GetInputVal).ToString()); Console.ReadLine(); } }   Now it gets really interesting because it looks like we have passed a method into a function in the main method by declaring… Console.WriteLine(myDelegate(GetInputVal).ToString()); So, what it the output? Well, try take a guess on what will happen – then copy the code and see if you got it right. Well that brings me to the end of this short explanation of Delegates. Hopefully it made sense!

    Read the article

  • Concurrency Utilities for Java EE Early Draft (JSR 236)

    - by arungupta
    Concurrency Utilities for Java EE is being worked as JSR 236 and has released an Early Draft. It provides concurrency capabilities to Java EE application components without compromising container integrity. Simple (common) and advanced concurrency patterns are easily supported without sacrificing usability. Using Java SE concurrency utilities such as java.util.concurrent API, java.lang.Thread and java.util.Timer in a Java EE application component such as EJB or Servlet are problematic since the container and server have no knowledge of these resources. JSR 236 enables concurrency largely by extending the Concurrency Utilities API developed under JSR-166. This also allows a consistency between Java SE and Java EE concurrency programming model. There are four main programming interfaces available: ManagedExecutorService ManagedScheduledExecutorService ContextService ManagedThreadFactory ManagedExecutorService is a managed version of java.util.concurrent.ExecutorService. The implementations of this interface are provided by the container and accessible using JNDI reference: <resource-env-ref>  <resource-env-ref-name>    concurrent/BatchExecutor  </resource-env-ref-name>  <resource-env-ref-type>    javax.enterprise.concurrent.ManagedExecutorService  </resource-env-ref-type><resource-env-ref> and available as: @Resource(name="concurrent/BatchExecutor")ManagedExecutorService executor; Its recommended to bind the JNDI references in the java:comp/env/concurrent subcontext. The asynchronous tasks that need to be executed need to implement java.lang.Runnable or java.util.concurrent.Callable interface as: public class MyTask implements Runnable { public void run() { // business logic goes here }} OR public class MyTask2 implements Callable<Date> {  public Date call() { // business logic goes here   }} The task is then submitted to the executor using one of the submit method that return a Future instance. The Future represents the result of the task and can also be used to check if the task is complete or wait for its completion. Future<String> future = executor.submit(new MyTask(), String.class);. . .String result = future.get(); Another example to submit tasks is: class MyTask implements Callback<Long> { . . . }class MyTask2 implements Callback<Date> { . . . }ArrayList<Callable> tasks = new ArrayList<();tasks.add(new MyTask());tasks.add(new MyTask2());List<Future<Object>> result = executor.invokeAll(tasks); The ManagedExecutorService may be configured for different properties such as: Hung Task Threshold: Time in milliseconds that a task can execute before it is considered hung Pool Info Core Size: Number of threads to keep alive Maximum Size: Maximum number of threads allowed in the pool Keep Alive: Time to allow threads to remain idle when # of threads > Core Size Work Queue Capacity: # of tasks that can be stored in inbound buffer Thread Use: Application intend to run short vs long-running tasks, accordingly pooled or daemon threads are picked ManagedScheduledExecutorService adds delay and periodic task running capabilities to ManagedExecutorService. The implementations of this interface are provided by the container and accessible using JNDI reference: <resource-env-ref>  <resource-env-ref-name>    concurrent/BatchExecutor  </resource-env-ref-name>  <resource-env-ref-type>    javax.enterprise.concurrent.ManagedExecutorService  </resource-env-ref-type><resource-env-ref> and available as: @Resource(name="concurrent/timedExecutor")ManagedExecutorService executor; And then the tasks are submitted using submit, invokeXXX or scheduleXXX methods. ScheduledFuture<?> future = executor.schedule(new MyTask(), 5, TimeUnit.SECONDS); This will create and execute a one-shot action that becomes enabled after 5 seconds of delay. More control is possible using one of the newly added methods: MyTaskListener implements ManagedTaskListener {  public void taskStarting(...) { . . . }  public void taskSubmitted(...) { . . . }  public void taskDone(...) { . . . }  public void taskAborted(...) { . . . } }ScheduledFuture<?> future = executor.schedule(new MyTask(), 5, TimeUnit.SECONDS, new MyTaskListener()); Here, ManagedTaskListener is used to monitor the state of a task's future. ManagedThreadFactory provides a method for creating threads for execution in a managed environment. A simple usage is: @Resource(name="concurrent/myThreadFactory")ManagedThreadFactory factory;. . .Thread thread = factory.newThread(new Runnable() { . . . }); concurrent/myThreadFactory is a JNDI resource. There is lot of interesting content in the Early Draft, download it, and read yourself. The implementation will be made available soon and also be integrated in GlassFish 4 as well. Some references for further exploring ... Javadoc Early Draft Specification concurrency-ee-spec.java.net [email protected]

    Read the article

  • Building Enterprise Smartphone App &ndash; Part 2: Platforms and Features

    - by Tim Murphy
    This is part 2 in a series of posts based on a talk I gave recently at the Chicago Information Technology Architects Group.  Feel free to leave feedback. In the previous post I discussed what reasons a company might have for creating a smartphone application.  In this installment I will cover some of history and state of the different platforms as well as features that can be leveraged for building enterprise smartphone applications. Platforms Before you start choosing a platform to develop your solutions on it is good to understand how we got here and what features you can leverage. History To my memory we owe all of this to a product called the Apple Newton that came out in 1987. It was the first PDA and back then I was much more of an Apple fan.  I was very impressed with this device even though it never really went anywhere.  The Palm Pilot by US Robotics was the next major advancement in PDA. It had a simple short hand window that allowed for quick stylus entry.. Later, Windows CE came out and started the broadening of the PDA market. After that it was the Palm and CE operating systems that started showing up on cell phones and for some time these were the two dominant operating systems that were distributed with devices from multiple hardware vendors. Current The iPhone was the first smartphone to take away the stylus and give us a multi-touch interface.  It was a revolution in usability and really changed the attractiveness of smartphones for the general public.  This brought us to the beginning of the current state of the market with the concept of an online store that makes it easy for customers to get new features and functionality on demand. With Android, Google made this more than a one horse race.  Not only did they come to compete, their low cost actually made them the leading OS.  Of course what made Android so attractive also is its major fault.  It is so open that it has been a target for malware which leaves consumers exposed.  Fortunately for Google though, most consumers aren’t aware of the threat that they are under. Although Microsoft had put out one of the first smart phone operating systems with CE it had to play catch up and finally came out with the Windows Phone.  They have gone for a market approach between those of iOS and Android.  They support multiple hardware vendors like Google, but they kept a certification process for applications that is similar to Apple.  They also created a user interface that was different enough to give it a clear separation from the other two platforms. The result of all this is hundreds of millions of smartphones being sold monthly across all three platforms giving us a wide range of choices and challenges when it comes to developing solutions. Features So what are the features that make these devices flexible enough be considered for use in the enterprise? The biggest advantage of today's devices is network connectivity.  The ability to access information from multiple sources at a moment’s notice is critical for businesses.  Add to that the ability to communicate over a variety of text, voice and video modes and we have a powerful starting point. Every smartphone has a cameras and they are not just useful for posting to Instagram. We are seeing more applications such as Bing vision that allow us to scan just about any printed code or text to find information.  These capabilities have been made available to developers in the form of standard libraries for reading barcodes of just about an flavor and optical character recognition (OCR) interpretation. Bluetooth give us the ability to communicate with multiple devices. Whether these are headsets, keyboard or printers the wireless communication capabilities are just starting to evolve.  The more these wireless communication protocols grow, the more opportunities we will see to transfer data between users and a variety of devices. Local storage of information that can be called up even when the device cannot reach the network is the other big capability.  This give users the ability to work offline as well and transmit information when connections are restored. These are the tools that we have to work with to build applications that can be leveraged to gain a competitive advantage for companies that implement them. Coming Up In the third installment I will cover key concerns that you face when building enterprise smartphone apps. del.icio.us Tags: smartphones,enterprise smartphone Apps,architecture,iOS,Android,Windows Phone

    Read the article

  • How to Hashtag (Without Being #Annoying)

    - by Mike Stiles
    The right tool in the wrong hands can be a dangerous thing. Giving a chimpanzee a chain saw would not be a pretty picture. And putting Twitter hashtags in the hands of social marketers who were never really sure how to use them can be equally unattractive. Boiled down, hashtags are for search and organization of tweets. A notch up from that, they can also be used as part of a marketing strategy. In terms of search, if you’re in the organic apple business, you want anyone who searches “organic” on Twitter to see your posts about your apples. It’s keyword tactics not unlike web site keyword search tactics. So get a clear idea of what keywords are relevant for your tweet. It’s reasonable to include #organic in your tweet. Is it fatal if you don’t hashtag the word? It depends on the person searching. If they search “organic,” your tweet’s going to come up even if you didn’t put the hashtag in front of it. If the searcher enters “#organic,” your tweet needs the hashtag. Err on the side of caution and hashtag it so it comes up no matter how the searcher enters it. You’ll also want to hashtag it for the second big reason people hashtag, organization. You can follow a hashtag. So can the rest of the Twitterverse. If you’re that into organic munchies, you can set up a stream populated only with tweets hashtagged #organic. If you’ve established a hashtag for your brand, like #nobugsprayapples, you (and everyone else) can watch what people are tweeting about your company. So what kind of hashtags should you include? They should be directly related to the core message of your tweet. Ancillary or very loosely-related hashtags = annoying. Hashtagging your brand makes sense. Hashtagging your core area of interest makes sense. Creating a specific event or campaign hashtag you want others to include and spread makes sense (the burden is on you to promote it and get it going). Hashtagging nearly every word in the tweet is highly annoying. Far and away, the majority of hashtagged words in such tweets have no relevance, are not terms that would be searched, and are not terms needed for categorization. It looks desperate and spammy. Two is fine. One is better. And it is possible to tweet with --gasp-- no hashtags! Make your hashtags as short as you can. In fact, if your brand’s name really is #nobugsprayapples, you’re burning up valuable, limited characters and risking the inability of others to retweet with added comments. Also try to narrow your topic hashtag down. You’ll find a lot of relevant users with #organic, but a lot of totally uninterested users with #food. Just as you can join online forums and gain credibility and a reputation by contributing regularly to that forum, you can follow hashtagged topics and gain the same kind of credibility in your area of expertise. Don’t just parachute in for the occasional marketing message. And if you’re constantly retweeting one particular person, stop it. It’s kissing up and it’s obvious. Which brings us to the king of hashtag annoyances, “hashjacking.” This is when you see what terms are hot and include them in your marketing tweet as a hashtag, even though it’s unrelated to your content. Justify it all you want, but #justinbieber has nothing to do with your organic apples. Equally annoying, piggybacking on a popular event’s hashtag to tweet something not connected to the event. You’re only fostering ill will and mistrust toward your account from the people you’ve tricked into seeing your tweet. Lastly, don’t @ mention people just to make sure they see your tweet. If the tweet’s not for them or about them, it’s spammy. What I haven’t covered is use of the hashtag for comedy’s sake. You’ll see this a lot and is a matter of personal taste. No one will search these hashtagged terms or need to categorize then, they’re just there for self-expression and laughs. Twitter is, after all, supposed to be fun.  What are some of your biggest Twitter pet peeves? #blogsovernow

    Read the article

  • What to leave when you're leaving

    - by BuckWoody
    There's already a post on this topic - sort of. I read this entry, where the author did a good job on a few steps, but I found that a few other tips might be useful, so if you want to check that one out and then this post, you might be able to put together your own plan for when you leave your job.  I once took over the system administrator (of which the Oracle and SQL Server servers were a part) at a mid-sized firm. The outgoing administrator had about a two- week-long scheduled overlap with me, but was angry at the company and told me "hey, I know this is going to be hard on you, but I want them to know how important I was. I'm not telling you where anything is or what the passwords are. Good luck!" He then quit that day. It took me about three days to find all of the servers and crack the passwords. Yes, the company tried to take legal action against the guy and all that, but he moved back to his home country and so largely got away with it. Obviously, this isn't the way to leave a job. Many of us have changed jobs in the past, and most of us try to be very professional about the transition to a new team, regardless of the feelings about a particular company. I've been treated badly at a firm, but that is no reason to leave a mess for someone else. So here's what you should put into place at a minimum before you go. Most of this is common sense - which of course isn't very common these days - and another good rule is just to ask yourself "what would I want to know"? The article I referenced at the top of this post focuses on a lot of documentation of the systems. I think that's fine, but in actuality, I really don't need that. Even with this kind of documentation, I still perform a full audit on the systems, so in the end I create my own system documentation. There are actually only four big items I need to know to get started with the systems: 1. Where is everything/everybody?The first thing I need to know is where all of the systems are. I mean not only the street address, but the closet or room, the rack number, the IU number in the rack, the SAN luns, all that. A picture here is worth a thousand words, which is why I really like Visio. It combines nice graphics, full text and all that. But use whatever you have to tell someone the physical locations of the boxes. Also, tell them the physical location of the folks in charge of those boxes (in case you aren't) or who share that responsibility. And by "where" in this case, I mean names and phones.  2. What do they do?For both the servers and the people, tell them what they do. If it's a database server, detail what each database does and what application goes to that, and who "owns" that application. In my mind, this is one of hte most important things a Data Professional needs to know. In the case of the other administrtors or co-owners, document each person's responsibilities.   3. What are the credentials?Logging on/in and gaining access to the buildings are things that the new Data Professional will need to do to successfully complete their job. This means service accounts, certificates, all of that. The first thing they should do, of course, is change the passwords on all that, but the first thing they need is the ability to do that!  4. What is out of the ordinary?This is the most tricky, and perhaps the next most important thing to know. Did you have to use a "special" driver for that video card on server X? Is the person that co-owns an application with you mentally unstable (like me) or have special needs, like "don't talk to Buck before he's had coffee. Nothing will make any sense"? Do you have service pack requirements for a specific setup? Write all that down. Anything that took you a day or longer to make work is probably a candidate here. This is my short list - anything you care to add? Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Deduping your redundancies

    - by nospam(at)example.com (Joerg Moellenkamp)
    Robin Harris of Storagemojo pointed to an interesting article about about deduplication and it's impact to the resiliency of your data against data corruption on ACM Queue. The problem in short: A considerable number of filesystems store important metadata at multiple locations. For example the ZFS rootblock is copied to three locations. Other filesystems have similar provisions to protect their metadata. However you can easily proof, that the rootblock pointer in the uberblock of ZFS for example is pointing to blocks with absolutely equal content in all three locatition (with zdb -uu and zdb -r). It has to be that way, because they are protected by the same checksum. A number of devices offer block level dedup, either as an option or as part of their inner workings. However when you store three identical blocks on them and the devices does block level dedup internally, the device may just deduplicated your redundant metadata to a block stored just once that is stored on the non-voilatile storage. When this block is corrupted, you have essentially three corrupted copies. Three hit with one bullet. This is indeed an interesting problem: A device doing deduplication doesn't know if a block is important or just a datablock. This is the reason why I like deduplication like it's done in ZFS. It's an integrated part and so important parts don't get deduplicated away. A disk accessed by a block level interface doesn't know anything about the importance of a block. A metadata block is nothing different to it's inner mechanism than a normal data block because there is no way to tell that this is important and that those redundancies aren't allowed to fall prey to some clever deduplication mechanism. Robin talks about this in regard of the Sandforce disk controllers who use a kind of dedup to reduce some of the nasty effects of writing data to flash, but the problem is much broader. However this is relevant whenever you are using a device with block level deduplication. It's just the point that you have to activate it for most implementation by command, whereas certain devices do this by default or by design and you don't know about it. However I'm not perfectly sure about that ? given that storage administration and server administration are often different groups with different business objectives I would ask your storage guys if they have activated dedup without telling somebody elase on their boxes in order to speak less often with the storage sales rep. The problem is even more interesting with ZFS. You may use ditto blocks to protect important data to store multiple copies of data in the pool to increase redundancy, even when your pool just consists out of one disk or just a striped set of disk. However when your device is doing dedup internally it may remove your redundancy before it hits the nonvolatile storage. You've won nothing. Just spend your disk quota on the the LUNs in the SAN and you make your disk admin happy because of the good dedup ratio However you can just fall in this specific "deduped ditto block"trap when your pool just consists out of a single device, because ZFS writes ditto blocks on different disks, when there is more than just one disk. Yet another reason why you should spend some extra-thought when putting your zpool on a single LUN, especially when the LUN is sliced and dices out of a large heap of storage devices by a storage controller. However I have one problem with the articles and their specific mention of ZFS: You can just hit by this problem when you are using the deduplicating device for the pool. However in the specifically mentioned case of SSD this isn't the usecase. Most implementations of SSD in conjunction with ZFS are hybrid storage pools and so rotating rust disk is used as pool and SSD are used as L2ARC/sZIL. And there it simply doesn't matter: When you really have to resort to the sZIL (your system went down, it doesn't matter of one block or several blocks are corrupt, you have to fail back to the last known good transaction group the device. On the other side, when a block in L2ARC is corrupt, you simply read it from the pool and in HSP implementations this is the already mentioned rust. In conjunction with ZFS this is more interesting when using a storage array, that is capable to do dedup and where you use LUNs for your pool. However as mentioned before, on those devices it's a user made decision to do so, and so it's less probable that you deduplicating your redundancies. Other filesystems lacking acapability similar to hybrid storage pools are more "haunted" by this problem of SSD using dedup-like mechanisms internally, because those filesystem really store the data on the the SSD instead of using it just as accelerating devices. However at the end Robin is correct: It's jet another point why protecting your data by creating redundancies by dispersing it several disks (by mirror or parity RAIDs) is really important. No dedup mechanism inside a device can dedup away your redundancy when you write it to a totally different and indepenent device.

    Read the article

  • If I were in a Silverlight focus group, here is ten things I would say.

    - by mbcrump
    Silverlight is a great product right off the shelf. I use it, love it and spend a lot of time helping the community understand it. This however, doesn’t mean that I don’t think that it can get better. If I were invited to a Microsoft Focus Group about Silverlight here is 10 things I would say:  We need more navigation templates. I’ve found (4) templates that Microsoft has released (Cosmo, Windows 7, Accent and JetPack). This number needs to be around 16. In order to get more people developing for Silverlight, we need to give them a variety of templates to get them off the ground quickly. Silverlight needs to ship with the next version of Windows. At least version 4 needs to be pre-installed on Windows going forward. It’s small, in its own sandbox and I cannot find a reason for it not to be included. Silverlight needs to run on more platforms.  iOS and Android are the key here. I think Microsoft should shoot for Android first since I believe Android will take the lead in the mobile market (at least for the short-term). It would also be great to see Microsoft use Silverlight as the focus on their new tablets / “AppleTV”. I would even invest in getting it working with Kinect. When creating a new project in Silverlight, we should have the option to create a Unit Test. Most Silverlight developers are not unit testing. If this is surprising to you then you need to get out and talk to more developers. I partially blame this on Microsoft. When you create a new ASP.NET MVC application, you simply put a check to create a Unit Test project. We need the same thing for Silverlight. We should steer the developer into the right direction. Design patterns such as MVVM need to be easier to implement in Silverlight solutions.  I’d go so far as to say that MVVM Light should ship with Visual Studio. With the project / item templates and code snippets, Laurent puts you into the right direction. This is the way that it should have been. Easy for the 9-5 developer to grasp. I believe the majority of developers use code behind because that’s what is in all the demos provided by Microsoft. They are not trying to write sucky code it is that they simply don’t know a better way.  The XAP Files should be obfuscated/unused references deleted by default when in “Release” mode. A better Silverlight experience starts with a smaller XAP file. The less that a user has to download is the better, even with the majority of people on broadband. I would also recommend built-in obfuscation by Microsoft. People are paranoid that they can rename the .zip and run it through reflector. Get rid of the boring install experiences. Here is a great write up on what I’m talking about. The default “Install Silverlight” and “Loading screens” suck. They suck bad. We need a choice of templates that a professional designer has created.  Silverlight needs to supports more image formats. For example: it would be great to use .gif’s without converting them to .png.    Switching between Blend 4 and VS2010 to develop a Silverlight application is a pain. Probably one of the biggest issues that I can’t think of a good solution for. It would be nice if VS2012 had the best of both worlds and you never have to leave VS. We need reporting controls with SSRS included with the Silverlight Toolkit. I can’t think of another control that we need built into the toolkit. It would also be helpful to have export to .xls, .pdf and .doc included with the control. I hope that this post will at least get a few people talking. Who knows, Microsoft could be working on these things right now. Thanks for reading!  Subscribe to my feed CodeProject

    Read the article

  • Using XA Transactions in Coherence-based Applications

    - by jpurdy
    While the costs of XA transactions are well known (e.g. increased data contention, higher latency, significant disk I/O for logging, availability challenges, etc.), in many cases they are the most attractive option for coordinating logical transactions across multiple resources. There are a few common approaches when integrating Coherence into applications via the use of an application server's transaction manager: Use of Coherence as a read-only cache, applying transactions to the underlying database (or any system of record) instead of the cache. Use of TransactionMap interface via the included resource adapter. Use of the new ACID transaction framework, introduced in Coherence 3.6.   Each of these may have significant drawbacks for certain workloads. Using Coherence as a read-only cache is the simplest option. In this approach, the application is responsible for managing both the database and the cache (either within the business logic or via application server hooks). This approach also tends to provide limited benefit for many workloads, particularly those workloads that either have queries (given the complexity of maintaining a fully cached data set in Coherence) or are not read-heavy (where the cost of managing the cache may outweigh the benefits of reading from it). All updates are made synchronously to the database, leaving it as both a source of latency as well as a potential bottleneck. This approach also prevents addressing "hot data" problems (when certain objects are updated by many concurrent transactions) since most database servers offer no facilities for explicitly controlling concurrent updates. Finally, this option tends to be a better fit for key-based access (rather than filter-based access such as queries) since this makes it easier to aggressively invalidate cache entries without worrying about when they will be reloaded. The advantage of this approach is that it allows strong data consistency as long as optimistic concurrency control is used to ensure that database updates are applied correctly regardless of whether the cache contains stale (or even dirty) data. Another benefit of this approach is that it avoids the limitations of Coherence's write-through caching implementation. TransactionMap is generally used when Coherence acts as system of record. TransactionMap is not generally compatible with write-through caching, so it will usually be either used to manage a standalone cache or when the cache is backed by a database via write-behind caching. TransactionMap has some restrictions that may limit its utility, the most significant being: The lock-based concurrency model is relatively inefficient and may introduce significant latency and contention. As an example, in a typical configuration, a transaction that updates 20 cache entries will require roughly 40ms just for lock management (assuming all locks are granted immediately, and excluding validation and writing which will require a similar amount of time). This may be partially mitigated by denormalizing (e.g. combining a parent object and its set of child objects into a single cache entry), at the cost of increasing false contention (e.g. transactions will conflict even when updating different child objects). If the client (application server JVM) fails during the commit phase, locks will be released immediately, and the transaction may be partially committed. In practice, this is usually not as bad as it may sound since the commit phase is usually very short (all locks having been previously acquired). Note that this vulnerability does not exist when a single NamedCache is used and all updates are confined to a single partition (generally implying the use of partition affinity). The unconventional TransactionMap API is cumbersome but manageable. Only a few methods are transactional, primarily get(), put() and remove(). The ACID transactions framework (accessed via the Connection class) provides atomicity guarantees by implementing the NamedCache interface, maintaining its own cache data and transaction logs inside a set of private partitioned caches. This feature may be used as either a local transactional resource or as logging XA resource. However, a lack of database integration precludes the use of this functionality for most applications. A side effect of this is that this feature has not seen significant adoption, meaning that any use of this is subject to the usual headaches associated with being an early adopter (greater chance of bugs and greater risk of hitting an unoptimized code path). As a result, for the moment, we generally recommend against using this feature. In summary, it is possible to use Coherence in XA-oriented applications, and several customers are doing this successfully, but it is not a core usage model for the product, so care should be taken before committing to this path. For most applications, the most robust solution is normally to use Coherence as a read-only cache of the underlying data resources, even if this prevents taking advantage of certain product features.

    Read the article

  • What are some concise and comprehensive introductory guide to unit testing for a self-taught programmer [closed]

    - by Superbest
    I don't have much formal training in programming and I have learned most things by looking up solutions on the internet to practical problems I have. There are some areas which I think would be valuable to learn, but which ended up both being difficult to learn and easy to avoid learning for a self-taught programmer. Unit testing is one of them. Specifically, I am interested in tests in and for C#/.NET applications using Microsoft.VisualStudio.TestTools in Visual Studio 2010 and/or 2012, but I really want a good introduction to the principles so language and IDE shouldn't matter much. At this time I'm interested in relatively trivial tests for small or medium sized programs (development time of weeks or months and mostly just myself developing). I don't necessarily intend to do test-driven development (I am aware that some say unit testing alone is supposed to be for developing features in TDD, and not an assurance that there are no bugs in the software, but unit testing is often the only kind of testing for which I have resources). I have found this tutorial which I feel gave me a decent idea of what unit tests and TDD looks like, but in trying to apply these ideas to my own projects, I often get confused by questions I can't answer and don't know how to answer, such as: What parts of my application and what sorts of things aren't necessarily worth testing? How fine grained should my tests be? Should they test every method and property separately, or work with a larger scope? What is a good naming convention for test methods? (since apparently the name of the method is the only way I will be able to tell from a glance at the test results table what works in my program and what doesn't) Is it bad to have many asserts in one test method? Since apparently VS2012 reports only that "an Assert.IsTrue failed within method MyTestMethod", and if MyTestMethod has 10 Assert.IsTrue statements, it will be irritating to figure out why a test is failing. If a lot of the functionality deals with writing and reading data to/from the disk in a not-exactly trivial fashion, how do I test that? If I provide a bunch of files as input by placing them in the program's directory, do I have to copy those files to the test project's bin/Debug folder now? If my program works with a large body of data and execution takes minutes or more, should my tests have it do the whole use all of the real data, a subset of it, or simulated data? If latter, how do I decide on the subset or how to simulate? Closely related to the previous point, if a class is such that its main operation happens in a state that is arrived to by the program after some involved operations (say, a class makes calculations on data derived from a few thousands of lines of code analyzing some raw data) how do I test just that class without inevitably ending up testing that class and all the other code that brings it to that state along with it? In general, what kind of approach should I use for test initialization? (hopefully that is the correct term, I mean preparing classes for testing by filling them in with appropriate data) How do I deal with private members? Do I just suck it up and assume that "not public = shouldn't be tested"? I have seen people suggest using private accessors and reflection, but these feel like clumsy and unsuited for regular use. Are these even good ideas? Is there anything like design patterns concerning testing specifically? I guess the main themes in what I'd like to learn more about are, (1) what are the overarching principles that should be followed (or at least considered) in every testing effort and (2) what are popular rules of thumb for writing tests. For example, at one point I recall hearing from someone that if a method is longer than 200 lines, it should be refactored - not a universally correct rule, but it has been quite helpful since I'd otherwise happily put hundreds of lines in single methods and then wonder why my code is so hard to read. Similarly I've found ReSharpers suggestions on member naming style and other things to be quite helpful in keeping my codebases sane. I see many resources both online and in print that talk about testing in the context of large applications (years of work, 10s of people or more). However, because I've never worked on such large projects, this context is very unfamiliar to me and makes the material difficult to follow and relate to my real world problems. Speaking of software development in general, advice given with the assumptions of large projects isn't always straightforward to apply to my own, smaller endeavors. Summary So my question is: What are some resources to learn about unit testing, for a hobbyist, self-taught programmer without much formal training? Ideally, I'm looking for a short and simple "bible of unit testing" which I can commit to memory, and then apply systematically by repeatedly asking myself "is this test following the bible of testing closely enough?" and then amending discrepancies if it doesn't.

    Read the article

  • "Mega Menus" for SEO [duplicate]

    - by Thought Space Designs
    This question already has an answer here: How do I handle having to many links on a webpage because of my menu 4 answers I'm using the term "Mega Menus" loosely here. I'm redesigning my WordPress site (it's going to be responsive), and as part of the redesign, I was debating incorporating some sort of descriptive menu setup. For example, normal navigation drop down menus come in the form of unordered lists of links like so: <nav> <ul> <li> <a href="#">Link1</a> </li> <li> <a href="#">Link2</a> </li> <li> <a href="#">Link3</a> <ul> <li> <a href="#">Sub Link1</a> </li> <li> <a href="#">Sub Link2</a> </li> <li> <a href="#">Sub Link3</a> </li> </ul> </li> <li> <a href="#">Link4</a> </li> </ul> </nav> What I'm looking to do is build my drop down menus with more information than your standard menu. For example, I have a top level link named "Team", and under that link, I want to make a large drop down that contains head shots, headers (in the form of styled p tags) and brief (<100 words) descriptions of each team member (only 2 currently). I want to accompany this with a "Read More" link that takes you to their actual team page. This is just one example, of course, and the other top level links would also have descriptive drop downs in the same fashion. On mobile, I was planning on hiding the "mega menu", and delivering a standard unordered list of links. Here's what I was thinking for overall structure and syntax: <nav> <ul> <li> <a href="#">Home</a> </li> <li> <a href="#">About</a> </li> <li> <a href="#">Team</a> <ul> <!-- DESKTOP --> <li class="mega-menu row"> <a class="col-sm-6" href="#"> <div class="row"> <div class="col-sm-4"> <img src="#" alt="Team Member 1" /> </div> <div class="col-sm-8"> <p class="header">Team Member 1</p> <p>Short description goes here.</p> </div> </div> </a> <a class="col-sm-6" href="#"> <!-- OTHER TEAM MEMBER INFO --> </a> </li> <!-- END DESKTOP --> <!-- MOBILE --> <li> <a href="#">Team Member 1</a> </li> <li> <a href="#">Team Member 2</a> </li> <!-- END MOBILE --> </ul> </li> <li> <a href="#">Contact</a> </li> </ul> </nav> Can anybody think of any potential SEO ramifications of doing this? I'm not going to be loading these menus full of links, so it shouldn't hurt page rank, but what are the effects of having a good bit of text and maybe even forms within nav elements? Is there such a thing as overloading nav with HTML? EDIT: Here's an example of what the menu would look like rendered on desktop. I'm currently hovering the "Team" menu, but you can't see because my mouse went away when I took the screenshot. EDIT 2: This question is not a duplicate. I'm not going to have "too many" links in my menus. I'm wondering how having images and text inside of header navigation will affect my menus. Also, I don't just want "yes, this is bad" answers. Please cite your sources and be specific with reasoning.

    Read the article

  • 11gR2 ??????????

    - by Allen Gao
    ???????11gR2 GI????????????????????,?10g????,???????GI?????????????1.Ocssd.bin:????????10g??????????,???????(Node Monitoring)????(Group Management)?????????????“??????????”????????2.Cssdagent.bin/Cssdmonitor.bin:?2????11gR2??????????????ocssd.bin??????(Local HeartBeat),??????1??????????????????ocssd.bin???????????,????????ocssd.bin????????,??????,???????????10g??oclsomon/oclsvmon(?????????????)?oprocd????,????11gR2???????—rebootless restart,?????????11.2.0.2????????????,????????????(????????)??????ocssd.bin?????,??????????????,??????????GI stack?????,??GI stack??????????(short disk I/O timeout)??graceful shutdown,????????,??,????????????????????????11gR2 ??????????????1.Ocssd.log2.Cssdagent ? cssdmonitor logs<GI_home>/log/<node_name>/agent/ohasd/oracssdagent_root/oracssdagent_root.log<GI_home>/log/<node_name>/agent/ohasd/oracssdmonitor_root_root/oracssdmonitor_root.log3.Cluster alert log<GI_home>/log/<node_name>/alert<node_name>.log4.OS log5.OSW ?? CHM ????,??????????????????1.???????????????????????????????,??????10g???????????????????????????GI alert log ??,?????node2?2012-08-15 16:30:06.554 [cssd(11011) ]CRS-1612:Network communication with node node1 (1) missing for 50% of timeout interval.  Removal of this node from cluster in 14.510 seconds2012-08-15 16:30:13.586 [cssd(11011) ]CRS-1611:Network communication with node node1 (1) missing for 75% of timeout interval.  Removal of this node from cluster in 7.470 seconds2012-08-15 16:30:18.606 [cssd(11011) ]CRS-1610:Network communication with node node1 (1) missing for 90% of timeout interval.  Removal of this node from cluster in 2.450 seconds2012-08-15 16:30:21.073 [cssd(11011) ]CRS-1632:Node node1 is being removed from the cluster in cluster incarnation 2363798322012-08-15 16:30:21.086 [cssd(11011) ]CRS-1601:CSSD Reconfiguration complete. Active nodes are node2 .?????????????node1?????????????????,???????, node2?? node1 ?????????node1 ???,???node1 ???????????????(????os log ??OSW ????),??node1 ???????node2??node1?????????,????node1??????????,???reconfiguration,????????????,????????????,?11.2.0.2??,??rebootless restart???,node eviction ????????GI stack??,????????????,???node2?node1?????????,node1?ocssd.bin??????(????ocssd.log??)??node1???????????????,??node1??????GI node eviction????2.???????????????,?????10g???????,???????????3.??ocssd.bin ????Cssdagent/Cssdmonitor.bin????????????,??????,????,????oracssdagent_root.log ?oracssdmonitor_root.log ????????2012-07-23 14:09:58.506: [ USRTHRD][1095805248] (:CLSN00111: )clsnproc_needreboot: Impending reboot at 75% of limit 28030; disk timeout 28030, network timeout 26380, last heartbeat from CSSD at epoch seconds 1343023777.410, 21091 milliseconds ago based on invariant clock 269251595; now polling at 100 ms……2012-07-23 14:10:02.704: [ USRTHRD][1095805248] (:CLSN00111: )clsnproc_needreboot: Impending reboot at 90% of limit 28030; disk timeout 28030, network timeout 26380, last heartbeat from CSSD at epoch seconds 1343023777.410, 25291 milliseconds ago based on invariant clock 269251595; now polling at 100 ms……???????????????timeout???28 ???(misscount – reboot time)?4.?????????????????? ??????????????????????,????ocssd.bin??????,?????????????,?????????????ocssd.bin??,????????os???????????OSW??,???? ??????? cpu ???Linux OSWbb v5.0 node1SNAP_INTERVAL 30CPU_COUNT 8OSWBB_ARCHIVE_DEST /osw/archiveprocs -----------memory---------- ---swap-- -----io---- -system-- -----cpu------r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id wa……zzz ***Mon Aug 30 17:55:21 CST 2012158  6 4200956 923940   7664 19088464    0    0  1296  3574 11153 231579  0 100  0  0  0zzz ***Mon Aug 30 17:55:53 CST 2012135  4 4200956 923760   7812 19089344    0    0     4    45  570 14563  0 100  0  0  0zzz ***Mon Aug 30 17:56:53 CST 2012126  2 4200956 923784   8396 19083620    0    0   196  1121  651 15941  2 98  0  0  0?????????????,????10g??????11gR2????????????????,??????,????????Note 1050693.1 : Troubleshooting 11.2 Clusterware Node Evictions (Reboots)

    Read the article

  • How do I stop and repair a RAID 5 array that has failed and has I/O pending?

    - by Ben Hymers
    The short version: I have a failed RAID 5 array which has a bunch of processes hung waiting on I/O operations on it; how can I recover from this? The long version: Yesterday I noticed Samba access was being very sporadic; accessing the server's shares from Windows would randomly lock up explorer completely after clicking on one or two directories. I assumed it was Windows being a pain and left it. Today the problem is the same, so I did a little digging; the first thing I noticed was that running ps aux | grep smbd gives a lot of lines like this: ben 969 0.0 0.2 96088 4128 ? D 18:21 0:00 smbd -F root 1708 0.0 0.2 93468 4748 ? Ss 18:44 0:00 smbd -F root 1711 0.0 0.0 93468 1364 ? S 18:44 0:00 smbd -F ben 3148 0.0 0.2 96052 4160 ? D Mar07 0:00 smbd -F ... There are a lot of processes stuck in the "D" state. Running ps aux | grep " D" shows up some other processes including my nightly backup script, all of which need to access the volume mounted on my RAID array at some point. After some googling, I found that it might be down to the RAID array failing, so I checked /proc/mdstat, which shows this: ben@jack:~$ cat /proc/mdstat Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10] md0 : active raid5 sdb1[3](F) sdc1[1] sdd1[2] 2930271872 blocks level 5, 64k chunk, algorithm 2 [3/2] [_UU] unused devices: <none> And running mdadm --detail /dev/md0 gives this: ben@jack:~$ sudo mdadm --detail /dev/md0 /dev/md0: Version : 00.90 Creation Time : Sat Oct 31 20:53:10 2009 Raid Level : raid5 Array Size : 2930271872 (2794.53 GiB 3000.60 GB) Used Dev Size : 1465135936 (1397.26 GiB 1500.30 GB) Raid Devices : 3 Total Devices : 3 Preferred Minor : 0 Persistence : Superblock is persistent Update Time : Mon Mar 7 03:06:35 2011 State : active, degraded Active Devices : 2 Working Devices : 2 Failed Devices : 1 Spare Devices : 0 Layout : left-symmetric Chunk Size : 64K UUID : f114711a:c770de54:c8276759:b34deaa0 Events : 0.208245 Number Major Minor RaidDevice State 3 8 17 0 faulty spare rebuilding /dev/sdb1 1 8 33 1 active sync /dev/sdc1 2 8 49 2 active sync /dev/sdd1 I believe this says that sdb1 has failed, and so the array is running with two drives out of three 'up'. Some advice I found said to check /var/log/messages for notices of failures, and sure enough there are plenty: ben@jack:~$ grep sdb /var/log/messages ... Mar 7 03:06:35 jack kernel: [4525155.384937] md/raid:md0: read error NOT corrected!! (sector 400644912 on sdb1). Mar 7 03:06:35 jack kernel: [4525155.389686] md/raid:md0: read error not correctable (sector 400644920 on sdb1). Mar 7 03:06:35 jack kernel: [4525155.389686] md/raid:md0: read error not correctable (sector 400644928 on sdb1). Mar 7 03:06:35 jack kernel: [4525155.389688] md/raid:md0: read error not correctable (sector 400644936 on sdb1). Mar 7 03:06:56 jack kernel: [4525176.231603] sd 0:0:1:0: [sdb] Unhandled sense code Mar 7 03:06:56 jack kernel: [4525176.231605] sd 0:0:1:0: [sdb] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE Mar 7 03:06:56 jack kernel: [4525176.231608] sd 0:0:1:0: [sdb] Sense Key : Medium Error [current] [descriptor] Mar 7 03:06:56 jack kernel: [4525176.231623] sd 0:0:1:0: [sdb] Add. Sense: Unrecovered read error - auto reallocate failed Mar 7 03:06:56 jack kernel: [4525176.231627] sd 0:0:1:0: [sdb] CDB: Read(10): 28 00 17 e1 5f bf 00 01 00 00 To me it is clear that device sdb has failed, and I need to stop the array, shutdown, replace it, reboot, then repair the array, bring it back up and mount the filesystem. I cannot hot-swap a replacement drive in, and don't want to leave the array running in a degraded state. I believe I am supposed to unmount the filesystem before stopping the array, but that is failing, and that is where I'm stuck now: ben@jack:~$ sudo umount /storage umount: /storage: device is busy. (In some cases useful info about processes that use the device is found by lsof(8) or fuser(1)) It is indeed busy; there are some 30 or 40 processes waiting on I/O. What should I do? Should I kill all these processes and try again? Is that a wise move when they are 'uninterruptable'? What would happen if I tried to reboot? Please let me know what you think I should do. And please ask if you need any extra information to diagnose the problem or to help!

    Read the article

  • Useful software for netbook?

    - by Moayad Mardini
    I'm looking for recommendations of good software that are particularly useful for netbooks. Software that run great on small screens and low CPU/RAM requirments. I'll start off with the following : Operating Systems: Ubuntu Netbook Remix. Easy Peasy: A fork of Ubuntu Netbook Remix that was once called UBuntu EEE. It isn't just for eeePCs though. Definitely worth a look if vanilla Netbook Remix isn't cutting it. (MarkM) Damn Small Linux (Source) Windows 7: With trimming the installation or compressing the Windows directory to fit on an 8GB SSD. (Will Eddins) nLite: A utility to install a lightweight version of Windows XP without the unnecessary components (like Media Player, Internet Explorer, Outlook Express, MSN Explorer, Messenger...). Utilites: TouchFreeze: To disable the touch pad while typing (Source) InSSIDer: Not only does it make it easier to find and keep a wireless connection, but it turns a netbook into the perfect mobile tool for troubleshooting wireless networks. (phenry) AltMove: Adds more functionality to your mouse for interacting with windows. (Rob) ASUS Font Resizer Utility and other tools by ASUS, specific to ASUS Eee PC series. Internet: Run FileZilla FTP client for a small screen : You can hide a lot of FileZilla's interface parts in the View menu, even the directory trees. Go into Settings = Interface and move the message log next to the transfer queue, if you haven't hidden them both or you want to see them. Select a theme with 16x16 icons. (Source) IDEs and Text Editors: Best lightweight IDE/Text Editor: A question on Stack Overflow that has many good suggestions of IDEs and general text editors for programmers. What’s a good linux C/C++ IDE for a low-res screen?: IDEs for Linux-powered netbooks. Online tools: Dropbox: Since the Netbook has limited disk space, you would like to use Cloud Apps like Dropbox and Ubuntu One so that you don't run out of space especially if you are on a holiday. Later when you go back to your desktop with big hard disk,you can take out the files from your dropbox repo. (Manish Sinha) Google products: like Docs, Calendar and Reader (aviraldg) Web sites and software lists: Netbookfiles.com: Netbook specific software downloads. Software Apps to Maximise your Netbook Battery Power: Netbooks are known for their portability. Not only are they small and lightweight but with their increased power efficiency, batteries can last much longer than conventional laptops. This also means you no longer have to carry a power adapter with you! Several brands emphasis the longevity of the battery as a strong selling point, and for those people who travel a lot, it sure is. Free Must-Have Netbook Apps: Finding software for netbooks can present challenges due to limited hard drive space, processor power, RAM, and screen real-estate. That doesn't mean you have to do without essential programs. The apps below cover all the bases -- entertainment, productivity, security, and communication -- without compromising on performance or usability. Best of all, they're free! Useful Netbook Software: With short battery lives and small resolution screens Netbooks, unlike many other computers on the market, could so with some specific software for their use. Now, not all of those I’ve found are specifically designed for Netbooks, but all are relevant. And they’re designed for Windows XP. The question is community wiki, so feel free to edit it. Updated, thank you all for suggestions.

    Read the article

  • Debian, 6rd tunnel, and connection troubles

    - by Chris B
    Long story short I am having issues with IPv6 using a 6rd tunnel with my ISP, charter business. They offer a 6rd tunnel that I think I have properly set up, but the server doesn’t reply to every ipv6 request. When the server has the network interfaces idle with no traffic for about 10 minutes, then IPv6 stops accepting inbound connections. to re-allow it, I must go into the server, and make it do a outbound ipv6 connection (normally a ping) to start it back up. Whats weird though i that if I run iptraf when its not working, it still shows a inbound ipv6 packet… the server is just not replying, and I can’t figure out why. Also, if I try to access my server over IPv6 from a house about 1 mile away on the same ISP, it is never able to connect. it always times out, but again the iptraf shows a ipv6 inbound packet. Again, it just does not reply. To test if my server is accessible through IPv6 I always have to use my vzw 4g phone (they use IPv6) or ipv6proxy dot net. Here is all of the configuration information my ISP gives on there tunnel server: 6rd Prefix = 2602:100::/32 Border Relay Address = 68.114.165.1 6rd prefix length = 32 IPv4 mask length = 0 Here is my /etc/network/interfaces for ipv6 (used x's to block real addresses) auto charterv6 iface charterv6 inet6 v4tunnel address 2602:100:189f:xxxx::1 netmask 32 ttl 64 gateway ::68.114.165.1 endpoint 68.114.165.1 local 24.159.218.xxx up ip link set mtu 1280 dev charterv6 here is my iptables config filter :INPUT DROP [0:0] :fail2ban-ssh – [0:0] :OUTPUT ACCEPT [0:0] :FORWARD DROP [0:0] :hold – [0:0] -A INPUT -p tcp -m tcp —dport 22 -j fail2ban-ssh -A INPUT -m state —state RELATED,ESTABLISHED -j ACCEPT -A INPUT -p tcp -m multiport -j ACCEPT —dports 80,443,25,465,110,995,143,993,587,465,22 -A INPUT -i lo -j ACCEPT -A INPUT -p tcp -m tcp —dport 10000 -j ACCEPT -A INPUT -p tcp -m tcp —dport 5900:5910 -j ACCEPT -A fail2ban-ssh -j RETURN -A INPUT -p icmp -j ACCEPT COMMIT and last here is my ip6tables firewall config filter :INPUT DROP [1653:339023] :FORWARD DROP [0:0] :OUTPUT ACCEPT [60141:13757903] :hold – [0:0] -A INPUT -m state —state RELATED,ESTABLISHED -j ACCEPT -A INPUT -p tcp -m multiport —dports 80,443,25,465,110,995,143,993,587,465,22 -j ACCEPT -A INPUT -i lo -j ACCEPT -A INPUT -p tcp -m tcp —dport 10000 -j ACCEPT -A INPUT -p tcp -m tcp —dport 5900:5910 -j ACCEPT -A INPUT -p ipv6-icmp -j ACCEPT COMMIT So Summary: 1.iptraf always shows IPv6 traffic, so its always making it to the server 2.server stops replying on ipv6 after no traffic for awhile (10 minutesish) until a outbound connection is made, then the process repeats. 3.server is NEVER accessable vi same ISP (yet iptraf still shows ipv6 request) Notes: When I try to access it from the same ISP from across town, even with iptables and ip6tables allowing ALL inbound traffic, this is what iptraf shows. IPv6 (92 bytes) from 97.92.18.xxx to 24.159.218.xxx on eth0 ICMP dest unrch (port) (120 bytes) from 24.159.218.xxx to 97.92.18.xxx on eth1 its strange, like its trying to forward to LAN? (eth1 is LAN, eth0 is WAN) even with the IPv6 address being set in the hosts file to the servers domain name. With iptables set up normally with the above configurations it only says this: IPv6 (100 bytes) from 97.92.18.xxx to 24.159.218.xxx on eth0 Im REALLY stuck on this, and any help would be GREATLY appreciated.

    Read the article

  • How can I automatically update Flash Player whenever a new version is released? [closed]

    - by user219950
    Summary: Flash Player Update Service doesn't run on a reliable schedule, and doesn't automatically download and apply updates when it does run. Given the importance of having an up-to-date version of Flash Player installed (for those of us who don't use Chrome with its built-in player), I would like to find a way to ensure that new updates are promptly detected and installed. What follows are the details of my efforts to solve this problem on my own... Appendix A: Flash Player Update Service OK, way back in Flash Player 11.2 (or so?) Adobe added the Flash Player Update Service (FlashPlayerUpdateService.exe), it was supposed to keep the Flash Player updated... Upon installation, FPUS is configured to run as a Windows Service, with Start Type set to Manual. A Scheduled Task (Adobe Flash Player Updater.job) is added to start this service every hour. So far, so good - this set-up avoids having a constantly-running service, but makes sure that the checks are run often enough to catch any updates quickly. Google's software updater is configured in a similar fashion, and that works just fine... ...And yet, when I checked the version of my installed Flash Player, I found it was 11.6.602.180, which, based on looking at the timestamps of the files in C:\Windows\System32\Macromed\Flash was last updated (or installed) on Tue, Mar 12, 2013 --- 3/12/13, 5:00:08pm. I made this observation on Thu, Apr 25, 2013 --- 4/25/13, 7:00:00pm, and upon checking Adobe's website found that the current version of Flash Player was 11.7.700.169. That's over a month since the last update, with a new one clearly available on the website but with no indication that the hourly check running on my machine has noticed it or has any intention of downloading it. Appendix B: running the Flash Player updater manually Once upon a time, running FlashUtil32_<version_Plugin.exe -update plugin would give you a window with an Install button; pressing it would download the installer for the current version (automatically, without opening a browser) and run it, then you'd click thru that installer & be done. It was manual, but it worked! Finding my current installation out of date (see Appendix A), I first tried this manual update process. However... Running FlashUtil32_<version_ActiveX.exe -update activex (in my case, that's FlashUtil32_11_6_602_180_ActiveX.exe -update activex) ...only presents a window with a Download button, clicking that Download button opens my browser to the URL https://get3.adobe.com/flashplayer/update/activex. Running FlashUtil32_<version_Plugin.exe -update plugin (in my case, that's FlashUtil32_11_6_602_180_Plugin.exe -update plugin) ...only presents a window with a Download button, clicking that Download button opens my browser to the URL https://get3.adobe.com/flashplayer/update/plugin. I could continue with the Download page it sent me to, uncheck the foistware box ("Free! McAfee Security Scan Plus"), download that installer (ActiveX, no foistware: install_flashplayer11x32axau_mssd_aih.exe, Plugin, no foistware: install_flashplayer11x32au_mssd_aih.exe) & probably have an updated Flash...but then, what is the point of the Flash Player Update Service if I have to manually download & run another exe? Epilogue I've since come to suspect that the update service is intentionally hobbled to drive early adopters to the manual download page. If this is true, there's probably no solution to this short of writing my own updater; hopefully I am wrong.

    Read the article

< Previous Page | 202 203 204 205 206 207 208 209 210 211 212 213  | Next Page >