Search Results

Search found 22641 results on 906 pages for 'use case'.

Page 241/906 | < Previous Page | 237 238 239 240 241 242 243 244 245 246 247 248  | Next Page >

  • Named arguments (parameters) as a readability aid

    - by Damian Mehers
    A long time ago I programmed a lot in ADA, and it was normal to name arguments when invoking a function - SomeObject.DoSomething(SomeParameterName = someValue); Now that C# supports named arguments, I'm thinking about reverting to this habit in situations where it might not be obvious what an argument means. You might argue that it should always be obvious what an argument means, but if you have a boolean argument, and callers are passing in "true" or "false" then qualifying the value with the name makes the call site more readable. contentFetcher.DownloadNote(note, manual : true); I guess I could create Enums instead of using true or false (Manual, Automatic in this case). What do you think about occasionally using named arguments to make code easier to read?

    Read the article

  • Should I just always Convert.ToInt32 my integers to account for potential nullable integers?

    - by Rowan Freeman
    If my MSSQL database contains a data type that is NULL (i.e. null is allowed) then ORMs, such as EntityFramework in my case, create .NET objects that are nullable. This is great, and the way I use nullables is like this: C# int? someInt = 5; int newInt = someInt.Value; // woot VB.NET Dim someInt As Integer? Dim newInt As Integer = someInt.Value ' hooray However, recently I had to make a change to the database to make an Id field no longer NULL (nullable). This means that .Value is now broken. This is a nuisance if the Id property is used a lot. One solution that I thought of is to just use Convert.ToInt32 on Id fields so it doesn't matter if an int is nullable or not. C# int newInt = Convert.ToInt32(someInt); // always compiles VB.NET Dim newInt As Integer = Convert.ToInt32(someInt) ' always compiles Is this a bad approach and are there any alternatives?

    Read the article

  • The SQL Server Community

    - by AllenMWhite
    In case you weren't aware of it, I absolutely love the SQL Server community. The people I've gotten to know have amazing knowledge, and they love sharing that knowledge with anyone who wants to learn. How can you not love that? It's inspiring and humbling all at the same time. There are a number of venues where the SQL Server community comes together. I'm including Twitter , the PASS Summit , the various SQL Saturday events, SQLBits , Tech Ed , and the local user groups. Each of us takes part in...(read more)

    Read the article

  • Oversizing images to produce better looking pages?

    - by Joannes Vermorel
    In the past, improper image resizing used to be a big no-no of web design (not mentioning improper compression format). Hence, for years I have been sticking to the policy where images (PNG or JPG) are resized on the server to match the resolution pixel-wise they will have with the rendered page. Now, recently, I hastily designed a HTML draft with oversized images, using inline CSS style such as width:123px and height:123px to resize the images. To my (slight) surprise, the page turned out to look much better that way. Indeed, with better screen resolution, some people (like me), tend to browse with some level of zoom (aka 125% or even 150% zoom), otherwise fonts are just too small on-screen. Then, if the image is strictly sized, the enlarged image appears blurry (pixel interpolation effect), but if the image is oversized the results is much better. Obviously, oversizing images is not an acceptable pattern if your website is intended for mobile browsing, but is there case where it would be considered as acceptable? Especially if the extra page weight is small anyway.

    Read the article

  • Optimal Database design regarding functionality of letting user share posts by other users

    - by codecool
    I want to implement functionality which let user share posts by other users similar to what Facebook and Google+ share button and twitter retweet. There are 2 choices: 1) I create duplicate copy of the post and have a column which keeps track of the original post id and makes clear this is a shared post. 2) I have a separate table shared post where I save the post id which is a foreign key to post id in post table. Talking in terms of programming basically I keep pointer to the original post in a separate table and when need to get post posted by user and also shared ones I do a left join on post and shared post table Post(post_id(PK), post_content, posted_by) SharedPost(post_id(FK to Post.post_id), sharing_user, sharedfrom(in case someone shares from non owners profile)) I am in favour of second choice but wanted to know the advice of experts out there? One thing more posts on my webapp will be more on the lines of facebook size not tweet size.

    Read the article

  • Complete deployment to AWS

    - by Ionut
    I'm trying to deploy a Java Application to AWS free tier. I need the following: RDS provider. Using MySQL client S3 service. This is required for the Lucene Index and image uploading SES service. I need to be able to send emails to new registered users. Namecheap is my domain provider. EC2 instance Elastic Beanstalk instance. I managed to create an EC2 instance, upload the WAR file and link it to the Namecheap domain. However I find it difficult to link the other instances to the current application. I find the documentation a little messy and I can't find the right way to do this. Can you provide a simple walk through to deployment for this use case? Thanks!

    Read the article

  • Which ATI driver version Ubuntu 10.04 LTS used?

    - by Jack
    I have a trouble with ATI driver with Ubuntu 12.04 LTS. When I install ATI driver in Ubuntu 12.04 via "Additional Driver", then I can't shutdown Ubuntu, it showed blank (black) screen and my laptop still run. Sometimes, my screen like that: http://www.youtube.com/watch?v=U9_iygesbBM But when I install Ubuntu 10.04 and install ATI driver via "Additional Driver", it's very good and no trouble I've seen. It's sweet but 10.04 is old and is supported to 4/2013. So I want to know why Ubuntu 10.04 works good better than 12.04 with ATI driver (in my case)?

    Read the article

  • NEW - Oracle Certifications and Documentation Available for Pre-Acquisition Sun/BEA IdM Products

    - by Irina
    If you have been looking for Oracle certification information or documentation for the pre-Acquisition Sun/BEA Identity Management products, you can now find them at the Certifications Central Hub.Use this Hub if you're looking for Sun Identity Management documentation, certified configurations for Waveset, Identity Analytics, OpenSSO, and more. Scroll down, below the bullets, to the bottom of the table to find: Of course, you can still find a great wealth of certification information for current products at this hub, as in the past. Be sure to check before you install! In case you haven't used this page before, notice that you can get to the documentation, certifications and downloads for IdM products by clicking on "Identity Management" in the leftmost pane. In the new screen, you will see each IdM product, along with tabs for Downloads, Documentation, Community, and Learn More. Let us know if you don't find what you are looking for. Happy Trails.

    Read the article

  • Spreadsheet or writing an application?

    - by Lenny222
    When would you keep simple to medium-complex personal calculations in a spread sheet (Excel etc) and when would you write a small program or script for it? For example when you want to calculate what size of mortgage you can afford to buy a house. I could create a spreadsheet and have a nice tabular representation. On the other hand, if i would write a small script in a nice language (in my case Haskell), i'd have the security of a nice type system, preventing typos etc. What are the pro/cons in your opinion?

    Read the article

  • ASP.NET Multi-Select Radio Buttons

    - by Ajarn Mark Caldwell
    “HERESY!” you say, “Radio buttons are for single-select items!  If you want multi-select, use checkboxes!”  Well, I would agree, and that is why I consider this a significant bug that ASP.NET developers need to be aware of.  Here’s the situation. If you use ASP:RadioButton controls on your WebForm, then you know that in order to get them to behave properly, that is, to define a group in which only one of them can be selected by the user, you use the Group attribute and set the same value on each one.  For example: 1: <asp:RadioButton runat="server" ID="rdo1" Group="GroupName" checked="true" /> 2: <asp:RadioButton runat="server" ID="rdo2" Group="GroupName" /> With this configuration, the controls will render to the browser as HTML Input / Type=radio tags and when the user selects one, the browser will automatically deselect the other one so that only one can be selected (checked) at any time. BUT, if you user server-side code to manipulate the Checked attribute of these controls, it is possible to set them both to believe that they are checked. 1: rdo2.Checked = true; // Does NOT change the Checked attribute of rdo1 to be false. As long as you remain in server-side code, the system will believe that both radio buttons are checked (you can verify this in the debugger).  Therefore, if you later have code that looks like this 1: if (rdo1.Checked) 2: { 3: DoSomething1(); 4: } 5: else 6: { 7: DoSomethingElse(); 8: } then it will always evaluate the condition to be true and take the first action.  The good news is that if you return to the client with multiple radio buttons checked, the browser tries to clean that up for you and make only one of them really checked.  It turns out that the last one on the screen wins, so in this case, you will in fact end up with rdo2 as checked, and if you then make a trip to the server to run the code above, it will appear to be working properly.  However, if your page initializes with rdo2 checked and in code you set rdo1 to checked also, then when you go back to the client, rdo2 will remain checked, again because it is the last one and the last one checked “wins”. And this gets even uglier if you ever set these radio buttons to be disabled.  In that case, although the client browser renders the radio buttons as though only one of them is checked the system actually retains the value of both of them as checked, and your next trip to the server will really frustrate you because the browser showed rdo2 as checked, but your DoSomething1() routine keeps getting executed. The following is sample code you can put into any WebForm to test this yourself. 1: <body> 2: <form id="form1" runat="server"> 3: <h1>Radio Button Test</h1> 4: <hr /> 5: <asp:Button runat="server" ID="cmdBlankPostback" Text="Blank Postback" /> 6: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 7: <asp:Button runat="server" ID="cmdEnable" Text="Enable All" OnClick="cmdEnable_Click" /> 8: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 9: <asp:Button runat="server" ID="cmdDisable" Text="Disable All" OnClick="cmdDisable_Click" /> 10: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 11: <asp:Button runat="server" ID="cmdTest" Text="Test" OnClick="cmdTest_Click" /> 12: <br /><br /><br /> 13: <asp:RadioButton ID="rdoG1R1" GroupName="Group1" runat="server" Text="Group 1 Radio 1" Checked="true" /><br /> 14: <asp:RadioButton ID="rdoG1R2" GroupName="Group1" runat="server" Text="Group 1 Radio 2" /><br /> 15: <asp:RadioButton ID="rdoG1R3" GroupName="Group1" runat="server" Text="Group 1 Radio 3" /><br /> 16: <hr /> 17: <asp:RadioButton ID="rdoG2R1" GroupName="Group2" runat="server" Text="Group 2 Radio 1" /><br /> 18: <asp:RadioButton ID="rdoG2R2" GroupName="Group2" runat="server" Text="Group 2 Radio 2" Checked="true" /><br /> 19:  20: </form> 21: </body> 1: protected void Page_Load(object sender, EventArgs e) 2: { 3:  4: } 5:  6: protected void cmdEnable_Click(object sender, EventArgs e) 7: { 8: rdoG1R1.Enabled = true; 9: rdoG1R2.Enabled = true; 10: rdoG1R3.Enabled = true; 11: rdoG2R1.Enabled = true; 12: rdoG2R2.Enabled = true; 13: } 14:  15: protected void cmdDisable_Click(object sender, EventArgs e) 16: { 17: rdoG1R1.Enabled = false; 18: rdoG1R2.Enabled = false; 19: rdoG1R3.Enabled = false; 20: rdoG2R1.Enabled = false; 21: rdoG2R2.Enabled = false; 22: } 23:  24: protected void cmdTest_Click(object sender, EventArgs e) 25: { 26: rdoG1R2.Checked = true; 27: rdoG2R1.Checked = true; 28: } 29: 30: protected void Page_PreRender(object sender, EventArgs e) 31: { 32:  33: } After you copy the markup and page-behind code into the appropriate files.  I recommend you set a breakpoint on Page_Load as well as cmdTest_Click, and add each of the radio button controls to the Watch list so that you can walk through the code and see exactly what is happening.  Use the Blank Postback button to cause a postback to the server so you can inspect things without making any changes. The moral of the story is: if you do server-side manipulation of the Checked status of RadioButton controls, then you need to set ALL of the controls in a group whenever you want to change one.

    Read the article

  • How would you code an AI engine to allow communication in any programming language?

    - by Tokyo Dan
    I developed a two-player iPhone board game. Computer players (AI) can either be local (in the game code) or remote running on a server. In the 2nd case, both client and server code are coded in Lua. On the server the actual AI code is separate from the TCP socket code and coroutine code (which spawns a separate instance of AI for each connecting client). I want to be able to further isolate the AI code so that that part can be a module coded by anyone in their language of choice. How can I do this? What tecniques/technology would enable communication between the Lua TCP socket/coroutine code and the AI module?

    Read the article

  • yield – Just yet another sexy c# keyword?

    - by George Mamaladze
    yield (see NSDN c# reference) operator came I guess with .NET 2.0 and I my feeling is that it’s not as wide used as it could (or should) be.   I am not going to talk here about necessarity and advantages of using iterator pattern when accessing custom sequences (just google it).   Let’s look at it from the clean code point of view. Let's see if it really helps us to keep our code understandable, reusable and testable.   Let’s say we want to iterate a tree and do something with it’s nodes, for instance calculate a sum of their values. So the most elegant way would be to build a recursive method performing a classic depth traversal returning the sum.           private int CalculateTreeSum(Node top)         {             int sumOfChildNodes = 0;             foreach (Node childNode in top.ChildNodes)             {                 sumOfChildNodes += CalculateTreeSum(childNode);             }             return top.Value + sumOfChildNodes;         }     “Do One Thing” Nevertheless it violates one of the most important rules “Do One Thing”. Our  method CalculateTreeSum does two things at the same time. It travels inside the tree and performs some computation – in this case calculates sum. Doing two things in one method is definitely a bad thing because of several reasons: ·          Understandability: Readability / refactoring ·          Reuseability: when overriding - no chance to override computation without copying iteration code and vice versa. ·          Testability: you are not able to test computation without constructing the tree and you are not able to test correctness of tree iteration.   I want to spend some more words on this last issue. How do you test the method CalculateTreeSum when it contains two in one: computation & iteration? The only chance is to construct a test tree and assert the result of the method call, in our case the sum against our expectation. And if the test fails you do not know wether was the computation algorithm wrong or was that the iteration? At the end to top it all off I tell you: according to Murphy’s Law the iteration will have a bug as well as the calculation. Both bugs in a combination will cause the sum to be accidentally exactly the same you expect and the test will PASS. J   Ok let’s use yield! That’s why it is generally a very good idea not to mix but isolate “things”. Ok let’s use yield!           private int CalculateTreeSumClean(Node top)         {             IEnumerable<Node> treeNodes = GetTreeNodes(top);             return CalculateSum(treeNodes);         }             private int CalculateSum(IEnumerable<Node> nodes)         {             int sumOfNodes = 0;             foreach (Node node in nodes)             {                 sumOfNodes += node.Value;             }             return sumOfNodes;         }           private IEnumerable<Node> GetTreeNodes(Node top)         {             yield return top;             foreach (Node childNode in top.ChildNodes)             {                 foreach (Node currentNode in GetTreeNodes(childNode))                 {                     yield return currentNode;                 }             }         }   Two methods does not know anything about each other. One contains calculation logic another jut the iteration logic. You can relpace the tree iteration algorithm from depth traversal to breath trevaersal or use stack or visitor pattern instead of recursion. This will not influence your calculation logic. And vice versa you can relace the sum with product or do whatever you want with node values, the calculateion algorithm is not aware of beeng working on some tree or graph.  How about not using yield? Now let’s ask the question – what if we do not have yield operator? The brief look at the generated code gives us an answer. The compiler generates a 150 lines long class to implement the iteration logic.       [CompilerGenerated]     private sealed class <GetTreeNodes>d__0 : IEnumerable<Node>, IEnumerable, IEnumerator<Node>, IEnumerator, IDisposable     {         ...        150 Lines of generated code        ...     }   Often we compromise code readability, cleanness, testability, etc. – to reduce number of classes, code lines, keystrokes and mouse clicks. This is the human nature - we are lazy. Knowing and using such a sexy construct like yield, allows us to be lazy, write very few lines of code and at the same time stay clean and do one thing in a method. That's why I generally welcome using staff like that.   Note: The above used recursive depth traversal algorithm is possibly the compact one but not the best one from the performance and memory utilization point of view. It was taken to emphasize on other primary aspects of this post.

    Read the article

  • Which hidden files and directories do I need?

    - by Sammy Black
    In a previous question, I explained my situation/plan: backing up home directory on external drive, reformatting laptop drive, installing 14.04, putting home directory back. (It hasn't happened yet because I can't seem to find the down time, in case things aren't working right away.) It occurred to me that maybe I don't want all of those hidden files and directories (e.g. .local/share/ubuntuone/syncdaemon/, .cache/google-chrome/, etc.) Just judging by the amount of time in copying, I can tell that some of these hidden directories are large. Question: Are there any hidden directories that I obviously don't need/want when I have the laptop running an updated distribution? Will they cause conflicts? (I plan on copying the backed-up directory tree back onto the laptop with the --no-clobber option.)

    Read the article

  • How can I create a symlink to the location that Ubuntu 10.10 mounts a CD?

    - by Michael Curran
    In Ubuntu 10.10, when I insert a CD or DVD into my optical drive, the system mounts the CD in a folder called /media/XYZ where XYZ is the disk's label. This has cause problems with Wine, as in order for an application to verify that an application's CD is present, Wine uses a symlinks to point to a mounted CD's folder. In this case, that folder must be /media/XYZ, but when using a different application, the folder would be different. I would like to know if there is a way to create a symlink that will always point to the mounted folder from a given /dev/cdrom* device, or how to force the system to always mount CDs to the same address (i.e. /media/cdrom).

    Read the article

  • All video players display black screen.

    - by Dennis
    I'm working in 10.04 Lucid. All my video players (Movie Player and VLC) and the preview windows in editors (OpenShot and Pitivi) will only display a black screen when playing a video. The sound is fine and the videos work fine on other computers. I have tried multiple formats from varying sources .MOV taken from old ffmpeg projects, .AVI straight from a camera, .MP4 using h.264 from OpenShot on another system, .OGV from a gtk-recordmydesktop session on this very computer. I even get a pure black screen in the viewer when starting a remote session in VNC. This box has a GeForce 8400 GS using the Nvidia drivers in case it may be a card problem or setting.

    Read the article

  • The practical cost of swapping effects

    - by sebf
    Hello, I use XNA for my projects and on those forums I sometimes see references to the fact that swapping an effect for a mesh has a relatively high cost, which surprises me as I thought to swap an effect was simply a case of copying the replacement shader program to the GPU along with appropriate parameters. I wondered if someone could explain exactly what is costly about this process? And put, if possible, 'relatively' into context? For example say I wanted to use a short shader to help with picking, I would: Change the effect on every object, calculting a unique color to identify it and providing it to the shader. Draw all the objects to a render target in memory. Get the color from the target and use it to look up the selected object. What portion of the total time taken to complete that process would be spent swapping the shaders? My instincts would say that rendering the scene again, no matter how simple the shader, would be an order of magnitude slower than any other part of the process so why all the concern over effects?

    Read the article

  • Making Class Diagram for MVC Pattern Project

    - by iMohammad
    I have a question about making a class diagram for an MVC based college senior project. If we have 2 actors of users in my system, lets say Undergrad and Graduate students are the children of abstract class called User. (Generalisation) Each actor has his own features. My question, in such case, do we need to have these two actors in separate classes which inherits from the abstract class User? even though, I'm going to implement them as roles using one Model called User Model ? I think you can see my confusion here. I code using MVC pattern, but I've never made a class diagram for this pattern. Thank you in advance!

    Read the article

  • what's wrong with this Lua code (creating text inside listener in Corona)

    - by Greg
    If you double/triple click on the myObject here the text does NOT disappear. Why is this not working when there are multiple events being fired? That is, are there actually multiple "text" objects, with some existing but no longer having a reference to them held by the local "myText" variable? Do I have to manually removeSelf() on the local "myText" field before assigning it another "display.newText(...)"? display.setStatusBar( display.HiddenStatusBar ) local myText local function hideMyText(event) print ("hideMyText") myText.isVisible = false end local function showTextListener(event) if event.phase == "began" then print("showTextListener") myText = display.newText("Hello World!", 0, 0, native.systemFont, 30) timer.performWithDelay(1000, hideMyText, 1 ) end end -- Display object to press to show text local myObject = display.newImage( "inventory_button.png", display.contentWidth/2, display.contentHeight/2) myObject:addEventListener("touch", showTextListener) Question 2 - Also why is it the case that if I add a line BEFORE "myText = ..." of: a) "if myText then myText:removeSelf() end" = THIS FIXES THINGS, whereas b) "if myText then myText=nil end" = DOES NOT FIX THINGS Interested in hearing how Lua works here re the answer...

    Read the article

  • FAQ: Creating a new LDOM domain

    - by Owen Allen
    I got a question about creating LDOM domains: "I have a Server Pool set up, and I need to create a secondary LDom domain on a machine in the pool. When I click on the machine, though, the 'create logical domain' command is grayed out. The machine still has available CPU threads and free RAM. What's going on?" This one has an easy answer. In a Server Pool, the Create Logical Domain action is under the pool's actions, rather than the individual machine's actions. This is because the Server Pool decides where to put the new domain based on the Server Pool's placement policy. So, in this case, you need to select the Server Pool in the Assets section, and then create the new domain from there.

    Read the article

  • MPL with Commercial Use Restrictions (And Other Questions)

    - by PythEch
    So basicly I want to use MPL 2.0 for my open source software but I also want to forbid commercial use. I'm not a legal expert, that's why I'm asking. Should I use dual-license (MPL + Modified BSD License)? Or what does sublicensing mean? If I wanted to license my project, I would include a notice to the header. What should I do if want to dual-license or sublicense? Also, is it OK to use nicknames as copyright owners? I am not able to distribute additional files (e.g LICENSE.txt, README.md etc) with the software simply because it is just a JS script. By open-source I mean not obfuscated JS code. So in this case, am I forced to use GPL to make redistribution of obfuscated work illegal? Thanks for reading, any help is appreciated, answering all of the questions is not essential.

    Read the article

  • Would like some help in understanding rendering geometry vs textures

    - by Anon
    So I was just pondering whether it is more taxing on the GPU to render geometry or a texture. What I'm trying to see is whether there is a huge difference in rendering two scenes with the same setup: Scene 1: Example Object: A dirt road (nothing else) Geometry: Detailed road, with all the bumps, cracks and so forth done in the mesh Scene 2: Example Object: A dirt road (nothing else) Geometry: A simple mesh, in a form of a road, but in this case maps and textures are simulating cracks, bumps, etc... So of these two, which one is likely to tax the hardware more? Or is it not a like for like comparison? What would be the best way of doing something like this? Go heavy on the textures? Or have a blend of both?

    Read the article

  • How do I give proper attribution when distributing my modified Ambiance theme?

    - by WarriorIng64
    I made a modified version of 12.04's Ambiance that uses a dark sidebar for Nautilus, and I would like to redistribute it via e.g. gnome-look.org. From the Launchpad page for the light-themes package, it says the themes are available under a Creative Commons BY-SA 3.0 license. The way I understand it, I can distribute my modified theme so long as I provide proper attribution for it and place it under the same license. In this case, who do I attribute as the author of the original theme and where/how should I display this attribution?

    Read the article

  • 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

  • Offline apt-get update to age of cache

    - by James Haigh
    I have a script to quickly upgrade a Live or fresh system from cached files on a flash drive. In essence, it looks like this: # *Code to remove and symlink /var/cache/apt/ if currently empty of packages.* sudo apt-get dist-upgrade # Quick offline cached upgrade; not limited by slow WANs. echo $'\nMake sure Internet is reachable and press enter for complete online upgrade.'; read sudo apt-get update sudo apt-get dist-upgrade # Complete online upgrade. The problem is that the ‘cached upgrade’ seems to ignore the cached pkgcache.bin and srcpkgcache.bin which is where I assume apt-get update stores its changes, so the upgrade completes as if the system is up-to-date. Useless. So in that case, I need some code to apt-get update to the age of the package cache on my flash drive. This code would be placed between the 1st and 2nd lines of the code above.

    Read the article

  • What should I call the process of converting an object to a string?

    - by shabbychef
    We are having a game of 'semantic football' in the office over this matter: I am writing a method for an object which will represent the object as a string. That string should be such that when typed (more likely, cut and pasted) into the interpreter window (I will keep the language name out of this for now), will produce an object which is, for our purposes, identical to the one upon which the method was called. There is a spirited discussion over the 'best' name for this method. The terms pickle, serialize, deflate, etc have been proposed. However, it seems that those terms assume some process for the de-pickling (unserialization, etc) that is not necessarily the language interpreter itself. That is, they do not specifically refer to the case where strings of valid code are produced. This is closer to a quine, but we are re-producing the object not the code, so this is not quite right. any suggestions?

    Read the article

< Previous Page | 237 238 239 240 241 242 243 244 245 246 247 248  | Next Page >