Search Results

Search found 65160 results on 2607 pages for 'compile time constant'.

Page 285/2607 | < Previous Page | 281 282 283 284 285 286 287 288 289 290 291 292  | Next Page >

  • Cannot execute Java program: UnsupportedClassVersionError

    - by Ricko Devian
    I have installed JDK 6, but I can't execute a Java program. For example, I have made test.java. I compile it with javac tes.java and there's no error when I compile it, but when I want to execute that program it always displays an error. I execute the Java program with java tes. Exception in thread "main" java.lang.UnsupportedClassVersionError: tes : Unsupported major.minor version 51.0 at java.lang.ClassLoader.defineClass1(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:634) at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:142) at java.net.URLClassLoader.defineClass(URLClassLoader.java:277) at java.net.URLClassLoader.access$000(URLClassLoader.java:73) at java.net.URLClassLoader$1.run(URLClassLoader.java:212) at java.security.AccessController.doPrivileged(Native Method) at java.net.URLClassLoader.findClass(URLClassLoader.java:205) at java.lang.ClassLoader.loadClass(ClassLoader.java:321) at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:294) at java.lang.ClassLoader.loadClass(ClassLoader.java:266) Could not find the main class: tes. Program will exit. My javac version is 1.7.0, my java version is 1.6.0. Here is my tes.java code: class tes{ public static void main(String[]args){ System.out.println("hello"); } }

    Read the article

  • Compiling kernal problem

    - by James
    Hi, I have a hp pavilion dm3t with intel HD graphics running ubuntu 10.10 64 bit. I'm trying to compile and install a patched kernel according to this, https://launchpad.net/~kamalmostafa/+archive/linux-kamal-mjgbacklight So I downloaded the tarball from here (linked to from the page above): http://kernel.ubuntu.com/git?p=kamal/ubuntu-maverick.git;a=shortlog;h=refs/heads/mjg-backlight I untar'd it to a directory, entered the directory and did: make defconfig which was successful, so I did: make which seemed to work fine until it gave these errors: ubuntu/ndiswrapper/iw_ndis.c:1966: error: unknown field ‘num_private’ specified in initializer ubuntu/ndiswrapper/iw_ndis.c:1966: warning: initialization makes pointer from integer without a cast ubuntu/ndiswrapper/iw_ndis.c:1967: error: unknown field ‘num_private_args’ specified in initializer ubuntu/ndiswrapper/iw_ndis.c:1967: warning: excess elements in struct initializer ubuntu/ndiswrapper/iw_ndis.c:1967: warning: (near initialization for ‘ndis_handler_def’) ubuntu/ndiswrapper/iw_ndis.c:1970: error: unknown field ‘private’ specified in initializer ubuntu/ndiswrapper/iw_ndis.c:1970: warning: initialization makes integer from pointer without a cast ubuntu/ndiswrapper/iw_ndis.c:1970: error: initializer element is not computable at load time ubuntu/ndiswrapper/iw_ndis.c:1970: error: (near initialization for ‘ndis_handler_def.num_standard’) ubuntu/ndiswrapper/iw_ndis.c:1971: error: unknown field ‘private_args’ specified in initializer ubuntu/ndiswrapper/iw_ndis.c:1971: warning: initialization from incompatible pointer type make[2]: *** [ubuntu/ndiswrapper/iw_ndis.o] Error 1 make[1]: *** [ubuntu/ndiswrapper] Error 2 make: *** [ubuntu] Error 2 How can I compile and install this kernel successfully? I'm new to this and would appreciate any help.

    Read the article

  • create a .deb Package from scripts or binaries

    - by tdeutsch
    I searched for a simple way to create .deb Packages for things which have no source code to compile (configs, shellscripts, proprietary software). This was quite a problem because most of the package tutorials are assuming you have a source tarball you want to compile. Then I've found this short tutorial (german). Afterwards, I created a small script to create a simple repository. Like this: rm /export/my-repository/repository/* cd /home/tdeutsch/deb-pkg for i in $(ls | grep my); do dpkg -b ./$i /export/my-repository/repository/$i.deb; done cd /export/avanon-repository/repository gpg --armor --export "My Package Signing Key" > PublicKey apt-ftparchive packages ./ | gzip > Packages.gz apt-ftparchive packages ./ > Packages apt-ftparchive release ./ > /tmp/Release.tmp; mv /tmp/Release.tmp Release gpg --output Release.gpg -ba Release I added the key to the apt keyring and included the source like this: deb http://my.default.com/my-repository/ ./ It looks like the repo itself is working well (I ran into some problems, to fix them I needed to add the Packages twice and make the temp-file workaround for the Release file). I also put some downloaded .deb into the repo, it looks like they are also working without problems. But my self created packages didn't... Wenn i do sudo apt-get update, they are causing errors like this: E: Problem parsing dependency Depends E: Error occurred while processing my-printerconf (NewVersion2) E: Problem with MergeList /var/lib/apt/lists/my.default.com_my-repository_._Packages E: The package lists or status file could not be parsed or opened. Has anyone an idea what I did wrong?

    Read the article

  • PerlRegEx vs RegularExpressionsCore Delphi Units

    - by Jan Goyvaerts
    The RegularExpressionsCore unit that is part of Delphi XE is based on the latest class-based PerlRegEx unit that I developed. Embarcadero only made a few changes to the unit. These changes are insignificant enough that code written for earlier versions of Delphi using the class-based PerlRegEx unit will work just the same with Delphi XE. The unit was renamed from PerlRegEx to RegularExpressionsCore. When migrating your code to Delphi XE, you can choose whether you want to use the new RegularExpressionsCore unit or continue using the PerlRegEx unit in your application. All you need to change is which unit you add to the uses clause in your own units. Indentation and line breaks in the code were changed to match the style used in the Delphi RTL and VCL code. This does not change the code, but makes it harder to diff the two units. Literal strings in the unit were separated into their own unit called RegularExpressionsConsts. These strings are only used for error messages that indicate bugs in your code. If your code uses TPerlRegEx correctly then the user should not see any of these strings. My code uses assertions to check for out of bounds parameters, while Embarcadero uses exceptions. Again, if you use TPerlRegEx correctly, you should never get any assertions or exceptions. The Compile method raises an exception if the regular expression is invalid in both my original TPerlRegEx component and Embarcadero’s version. If your code allows the user to provide the regular expression, you should explicitly call Compile and catch any exceptions it raises so you can tell the user there is a problem with the regular expression. Even with user-provided regular expressions, you shouldn’t get any other assertions or exceptions if your code is correct. Note that Embarcadero owns all the rights to their RegularExpressionsCore unit. Like all the other RTL and VCL units, this unit cannot be distributed by myself or anyone other than Embarcadero. I do retain the rights to my original PerlRegEx unit which I will continue to make available for those using older versions of Delphi.

    Read the article

  • How do I initialize a Scala map with more than 4 initial elements in Java?

    - by GlenPeterson
    For 4 or fewer elements, something like this works (or at least compiles): import scala.collection.immutable.Map; Map<String,String> HAI_MAP = new Map4<>("Hello", "World", "Happy", "Birthday", "Merry", "XMas", "Bye", "For Now"); For a 5th element I could do this: Map<String,String> b = HAI_MAP.$plus(new Tuple2<>("Later", "Aligator")); But I want to know how to initialize an immutable map with 5 or more elements and I'm flailing in Type-hell. Partial Solution I thought I'd figure this out quickly by compiling what I wanted in Scala, then decompiling the resultant class files. Here's the scala: object JavaMapTest { def main(args: Array[String]) = { val HAI_MAP = Map(("Hello", "World"), ("Happy", "Birthday"), ("Merry", "XMas"), ("Bye", "For Now"), ("Later", "Aligator")) println("My map is: " + HAI_MAP) } } But the decompiler gave me something that has two periods in a row and thus won't compile (I don't think this is valid Java): scala.collection.immutable.Map HAI_MAP = (scala.collection.immutable.Map) scala.Predef..MODULE$.Map().apply(scala.Predef..MODULE$.wrapRefArray( scala.Predef.wrapRefArray( (Object[])new Tuple2[] { new Tuple2("Hello", "World"), new Tuple2("Happy", "Birthday"), new Tuple2("Merry", "XMas"), new Tuple2("Bye", "For Now"), new Tuple2("Later", "Aligator") })); I'm really baffled by the two periods in this: scala.Predef..MODULE$ I asked about it on #java on Freenode and they said the .. looked like a decompiler bug. It doesn't seem to want to compile, so I think they are probably right. I'm running into it when I try to browse interfaces in IntelliJ and am just generally lost. Based on my experimentation, the following is valid: Tuple2[] x = new Tuple2[] { new Tuple2<String,String>("Hello", "World"), new Tuple2<String,String>("Happy", "Birthday"), new Tuple2<String,String>("Merry", "XMas"), new Tuple2<String,String>("Bye", "For Now"), new Tuple2<String,String>("Later", "Aligator") }; scala.collection.mutable.WrappedArray<Tuple2> y = scala.Predef.wrapRefArray(x); There is even a WrappedArray.toMap() method but the types of the signature are complicated and I'm running into the double-period problem there too when I try to research the interfaces from Java.

    Read the article

  • About shared (static) Members and its behavior

    - by Allende
    I just realized that I can access shared members from instances of classes (probably this is not correct, but compile and run), and also learn/discover that, I can modify shared members, then create a new instance and access the new value of the shared member. My question is, what happens to the shared members, when it comes back to the "default" value (class declaration), how dangerous is it do this ? is it totally bad ? is it valid in some cases ?. If you want to test my point here is the code (console project vb.net) that I used to test shared members, as you can see/compile/run, the shared member "x" of the class "Hello" has default value string "Default", but at runtime it changes it, and after creating a new object of that class, this object has the new value of the shared member. Module Module1 Public Class hello Public Shared x As String = "Default" Public Sub New() End Sub End Class Sub Main() Console.WriteLine("hello.x=" & hello.x) Dim obj As New hello() Console.WriteLine("obj.x=" & obj.x) obj.x = "Default shared memeber, modified in object" Console.WriteLine("obj.x=" & obj.x) hello.x = "Defaul shared member, modified in class" Console.WriteLine("hello.x=" & hello.x) Dim obj2 As New hello() Console.WriteLine("obj2.x=" & obj2.x) Console.ReadLine() End Sub End Module UPDATE: First at all, thanks to everyone, each answer give feedback, I suppose, by respect I should choose one as "the answer", I don't want to be offensive to anyone, so please don't take it so bad if I didn't choose you answer.

    Read the article

  • Languages with a clear distinction between subroutines that are purely functional, mutating, state-changing, etc?

    - by CPX
    Lately I've become more and more frustrated that in most modern programming languages I've worked with (C/C++, C#, F#, Ruby, Python, JS and more) there is very little, if any, language support for determining what a subroutine will actually do. Consider the following simple pseudo-code: var x = DoSomethingWith(y); How do I determine what the call to DoSomethingWith(y) will actually do? Will it mutate y, or will it return a copy of y? Does it depend on global or local state, or is it only dependent on y? Will it change the global or local state? How does closure affect the outcome of the call? In all languages I've encountered, almost none of these questions can be answered by merely looking at the signature of the subroutine, and there is almost never any compile-time or run-time support either. Usually, the only way is to put your trust in the author of the API, and hope that the documentation and/or naming conventions reveal what the subroutine will actually do. My question is this: Does there exist any languages today that make symbolic distinctions between these types of scenarios, and places compile-time constraints on what code you can actually write? (There is of course some support for this in most modern languages, such as different levels of scope and closure, the separation between static and instance code, lambda functions, et cetera. But too often these seem to come into conflict with each other. For instance, a lambda function will usually either be purely functional, and simply return a value based on input parameters, or mutate the input parameters in some way. But it is usually possible to access static variables from a lambda function, which in turn can give you access to instance variables, and then it all breaks apart.)

    Read the article

  • Why is my shadowmap all white?

    - by Berend
    I was trying out a shadowmap. But all my shadow is white. I think there is some problem with my homogeneous component. Can anybody help me? The rest of my code is written in xna Here is the hlsl code I used float4x4 xWorld; float4x4 xView; float4x4 xProjection; struct VertexToPixel { float4 Position : POSITION; float4 ScreenPos : TEXCOORD1; float Depth : TEXCOORD2; }; struct PixelToFrame { float4 Color : COLOR0; }; //------- Technique: ShadowMap -------- VertexToPixel MyVertexShader(float4 inPos: POSITION0, float3 inNormal: NORMAL0) { VertexToPixel Output = (VertexToPixel)0; float4x4 preViewProjection = mul(xView, xProjection); float4x4 preWorldViewProjection = mul(xWorld, preViewProjection); Output.Position =mul(inPos, mul(xWorld, preViewProjection)); Output.Depth = Output.Position.z / Output.Position.w; Output.ScreenPos = Output.Position; return Output; } float4 MyPixelShader(VertexToPixel PSIn) : COLOR0 { PixelToFrame Output = (PixelToFrame)0; Output.Color = PSIn.ScreenPos.z/PSIn.ScreenPos.w; return Output.Color; } technique ShadowMap { pass Pass0 { VertexShader = compile vs_2_0 MyVertexShader(); PixelShader = compile ps_2_0 MyPixelShader(); } }

    Read the article

  • C++ Succinctly now available!

    - by Michael B. McLaughlin
    Over the summer I worked with SyncFusion to create an eBook based off of my C# to C++ guide for their free Succinctly Series of eBooks. Today the result, C++ Succinctly, was published for download. It is a free (registration required; they make tools and libraries for .NET development so you might get an occasional email from them – I’ve been signed up for a few months and have had maybe 3 emails total so it’s not horrible super spam or anything ) and you can download it as a PDF or a Kindle .MOBI file (or both). I’m excited with how it turned out and enjoyed working with the people at SyncFusion. The book contains a total of 20 code samples, which you can download from BitBucket (there’s a link very early in the book). Almost all of the code is also inline in the book itself so that you don’t need to worry about flipping back and forth between your dev machine and your eReader (but if you want to try to understand a concept better, you can easily download the code, open it up in VS 2012, and play around with it to see what happens when you tinker with things). The code does require Visual Studio 2012 because of its expanded support for C++11 features and since I wrote all of the samples as Console programs for clarity and compactness, you will need a version that supports C++ desktop development (currently VS 2012 Pro, Premium, or Ultimate). Sometime this Fall, Microsoft will be releasing Visual Studio 2012 Express for Windows Desktop which should provide a free way to use the samples. That said, I tested all of the samples with MinGW and only the StorageDurationSample will not compile with it due to the thread-local storage code. If you comment that out then you can compile and run all the samples with MinGW (or using a recent version of GCC in a GNU/Linux environment, or any other C++ compiler that provides the same level of C++11 support that Visual Studio 2012 does). I hope it proves helpful to those of you who choose to check it out!

    Read the article

  • ubuntu 10.04: boot error for custom compiled kernel - gave up wating for root device

    - by atharva
    Hi, I have installed lucid on my Lenevo Laptop (Y 410 series , x86 platoform) and it is working fine. Now I have compiled kernel 2.6.37 from the downloaded from the kernel tree. I followed usual procedure of compileing kernel (make menuconfig,make. make modules etc). Then I created the initrd image using mkinitramfs and updated my grub using upadate grub command. Update-grub detects the initrd image of the compiled kernel. However when I boot from this kernel it gives me following error: Gave up waiting for root device. Common problems: -Boot args (cat /proc/cmdline) -Check rootdelay= (did the system wait long enough?) -Check root= (did the system wait for the right device?) -Missing modules (cat /proc/modules; ls /dev) ALERT! root=UUID=/... does not exist and then it falls onto initramfs prompt. I have tried following solutions discussed in different ubuntu forums: 1. disable uuid and point root=/dev/sda8 (sda8 is where my kernele image resides (both default kernel and compiled one) from /etc/default/grub 2. compile kernel using CONFIG_DEVTMPFS=y suggested here Still I am unable to boot from the compile kernel. Could someone please suggest me the solution ?

    Read the article

  • Run Grunt task in Visual Studio Release Build with a bat file

    - by Aligned
    Originally posted on: http://geekswithblogs.net/Aligned/archive/2014/08/19/run-grunt-task-in-visual-studio-release-build-with-a.aspx 1. Add a BeforeBuild in your csproj file. Edit the xml with a text editor. <Target Name="BeforeBuild"> <Exec Condition="'$(Configuration)' == 'Release'" Command="script-optimize.bat" /> </Target> 2. Create the script-optimize.batREM "%~dp0" maps to the directory where this file exists cd %~dp0\..\YourProjectFolder call npm uninstall grunt call npm uninstall grunt call npm install --cache-min 604800 -g grunt-cli call npm install --cache-min 604800 grunt typescript requirejs copy less:compile less:mincompileThis grunt command will compile typescript, run the requireJs optimizer, complie and minimize less.3. Make it use the minified code when the Web.config compilation debug is set to false <!-- These CustomCollectFiles actions are used so that the Scripts-Release folder/files are included        when publishing even though they are not project references -->  <Target Name="CustomCollectFiles">    <ItemGroup>      <_CustomFiles Include="Scripts-Release\**\*" />  </ItemGroup>  </Target> That should be all you need to get a Grunt task to minify and combine JS (plus other tasks) in Visual Studio Release build with debug = false. This is a great video of Steve Sanderson talking about SPAs, npm, Knockout, Grunt, Gulp, ect. I highly recommend it.

    Read the article

  • Ubuntu 10.04: boot error for custom compiled kernel - gave up waiting for root device

    - by atharva
    I have installed lucid on my Lenevo Laptop (Y 410 series , x86 platform) and it is working fine. Now I have compiled kernel 2.6.37 downloaded from the kernel tree. I followed usual procedure of compiling kernel (make menuconfig, make, make modules etc). Then I created the initrd image using mkinitramfs and updated my grub using update-grub command. update-grub detects the initrd image of the compiled kernel. However when I boot from this kernel it gives me following error: Gave up waiting for root device. Common problems: -Boot args (cat /proc/cmdline) -Check rootdelay= (did the system wait long enough?) -Check root= (did the system wait for the right device?) -Missing modules (cat /proc/modules; ls /dev) ALERT! root=UUID=/... does not exist and then it falls onto initramfs prompt. I have tried following solutions discussed in different Ubuntu forums: disable uuid and point root=/dev/sda8 (sda8 is where my kernel image resides (both default kernel and compiled one) from /etc/default/grub compile kernel using CONFIG_DEVTMPFS=y suggested here Still I am unable to boot from the compile kernel. Could someone please suggest me the solution?

    Read the article

  • Interface (contract), Generics (universality), and extension methods (ease of use). Is it a right design?

    - by Saeed Neamati
    I'm trying to design a simple conversion framework based on these requirements: All developers should follow a predefined set of rules to convert from the source entity to the target entity Some overall policies should be able to be applied in a central place, without interference with developers' code Both the creation of converters and usage of converter classes should be easy To solve these problems in C# language, A thought came to my mind. I'm writing it here, though it doesn't compile at all. But let's assume that C# compiles this code: I'll create a generic interface called IConverter public interface IConverter<TSource, TTarget> where TSource : class, new() where TTarget : class, new() { TTarget Convert(TSource source); List<TTarget> Convert(List<TSource> sourceItems); } Developers would implement this interface to create converters. For example: public class PhoneToCommunicationChannelConverter : IConverter<Phone, CommunicationChannle> { public CommunicationChannel Convert(Phone phone) { // conversion logic } public List<CommunicationChannel> Convert(List<Phone> phones) { // conversion logic } } And to make the usage of this conversion class easier, imagine that we add static and this keywords to methods to turn them into Extension Methods, and use them this way: List<Phone> phones = GetPhones(); List<CommunicationChannel> channels = phones.Convert(); However, this doesn't even compile. With those requirements, I can think of some other designs, but they each lack an aspect. Either the implementation would become more difficult or chaotic and out of control, or the usage would become truly hard. Is this design right at all? What alternatives I might have to achieve those requirements?

    Read the article

  • Compilation problem [closed]

    - by Misery
    I am trying to compile a program called Triangle Mesh Generator. It is an open source code written in ANSI C. I need it to bo a callable lib so I am using a switch (#define) created by it's Author. However I get this error: /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../x86_64-linux-gnu/crt1.o: In function `_start': collect2: ld returned 1 exit status make: *** [triangle] Error 1 I am not sure why such an error occurs. Worth adding is that compiling it as a stand alone program is succesful. I am using GCC on 12.04. The lines quoted above constitute the total output. What I put in the terminal is just make in the proper folder. There are no other errors, warnings, or other messages. Link to the sources I found some additional instructions. I'll get back after reading them :] EDIT: I have found some additional instructions that let me compile it. Thanks for help. I am closing question right now. Regards

    Read the article

  • VMware Player 5.0 or VMware Workstation 9.0 after upgrade to Ubuntu 12.10

    The upgrade process Upgrading Ubuntu 12.04 to latest version 12.10 - aka Quantal Quetzal - is straight forward and you only need to follow the offical upgrade instructions. Short version on the console looks like this: sudo do-release-upgrade This will update the repository entries, and start the upgrade process. After some minutes or hours of download and installation, you have to reboot your system once to get the new kernel loaded. As time of writing, I'm on '3.5.0-17-generic'. And as with any modification of the kernel version, you have to compile the necessary kernel modules to get VMware Player or Workstation up and running. Usually, this happens the first time you try start your VMware software and that's it. Well, again not so this time. Getting the kernel patch Luckily, the community over VMware is very active and you can get a new kernel patch in the online forums here. Get the download and put in a folder have write permissions. Then you extract the archive on the console like so: tar -xjvf vmware9_kernel35_patch.tar.bz2 Then you change into the newly created folder: cd vmware9_kernel3.5_patch/ And you execute the available shell script as root (superuser) like so: sudo ./patch-modules_3.5.0.sh This will stop any running instances of VMware software, patches the source files and runs the compile process for your active environment. This might take some time depending on your machine, and once completed you can start VMware Player or Workstation as previously. In case that you are going to apply the patch again, the script will simply quit with the following output: /usr/lib/vmware/modules/source/.patched found. You have already patched your sources. Exiting You might remove the .patched file in case that you upgraded/changed your kernel and you need to apply the patch again. Disclaimer: The patch is "as-is" and the patcher is originally created by Artem S. Tashkinov, and later modified by An_tony. Please refer to the VMware forum in case of questions or problems. There are also patches available for older versions of VMware Player or Workstation.

    Read the article

  • Is there a better way to organize my module tests that avoids an explosion of new source files?

    - by luser droog
    I've got a neat (so I thought) way of having each of my modules produce a unit-test executable if compiled with the -DTESTMODULE flag. This flag guards a main() function that can access all static data and functions in the module, without #including a C file. From the README: -- Modules -- The various modules were written and tested separately before being coupled together to achieve the necessary basic functionality. Each module retains its unit-test, its main() function, guarded by #ifdef TESTMODULE. `make test` will compile and execute all the unit tests, producing copious output, but importantly exitting with an appropriate success or failure code, so the `make test` command will fail if any of the tests fail. Module TOC __________ test obj src header structures CONSTANTS ---- --- --- --- -------------------- m m.o m.c m.h mfile mtab TABSZ s s.o s.c s.h stack STACKSEGSZ v v.o v.c v.h saverec_ f.o f.c f.h file ob ob.o ob.c ob.h object ar ar.o ar.c ar.h array st st.o st.c st.h string di di.o di.c di.h dichead dictionary nm nm.o nm.c nm.h name gc gc.o gc.c gc.h garbage collector itp itp.c itp.h context osunix.o osunix.c osunix.h unix-dependent functions It's compile by a tricky bit of makefile, m:m.c ob.h ob.o err.o $(CORE) itp.o $(OP) cc $(CFLAGS) -DTESTMODULE $(LDLIBS) -o $@ $< err.o ob.o s.o ar.o st.o v.o di.o gc.o nm.o itp.o $(OP) f.o where the module is compiled with its own C file plus every other object file except itself. But it's creating difficulties for the kindly programmer who offered to write the Autotools files for me. So the obvious way to make it "less weird" would be to bust-out all the main functions into separate source files. But, but ... Do I gotta?

    Read the article

  • What's is the point of PImpl pattern while we can use interface for same purpose in C++?

    - by ZijingWu
    I see a lot of source code which using PIMPL idiom in C++. I assume Its purposes are hidden the private data/type/implementation, so it can resolve dependence, and then reduce compile time and header include issue. But interface class in C++ also have this capability, it can also used to hidden data/type and implementation. And to hidden let the caller just see the interface when create object, we can add an factory method in it declaration in interface header. The comparison is: Cost: The interface way cost is lower, because you doesn't even need to repeat the public wrapper function implementation void Bar::doWork() { return m_impl->doWork(); }, you just need to define the signature in the interface. Well understand: The interface technology is more well understand by every C++ developer. Performance: Interface way performance not worse than PIMPL idiom, both an extra memory access. I assume the performance is same. Following is the pseudocode code to illustrate my question: // Forward declaration can help you avoid include BarImpl header, and those included in BarImpl header. class BarImpl; class Bar { public: // public functions void doWork(); private: // You doesn't need to compile Bar.cpp after change the implementation in BarImpl.cpp BarImpl* m_impl; }; The same purpose can be implement using interface: // Bar.h class IBar { public: virtual ~IBar(){} // public functions virtual void doWork() = 0; }; // to only expose the interface instead of class name to caller IBar* createObject(); So what's the point of PIMPL?

    Read the article

  • What is a suitable way to correct this program? [closed]

    - by lamwaiman1988
    Possible Duplicate: How to deal with huge changes to a data specification? As described in this question, I need to modified a 4800 line c program to fulfill the new functional specification. However I have no idea of HOW to do it. Background: There is a updated version of a header file which the program depends on. I've speculated the changes and found that at least 30 critical changes are present. With the new header file, the c program cannot even compile. The error is too much that the compile said "too much error" and don't list every one. Unlike normal modification which can be easily test, this has no way to test without overcoming the 30 critical changes. For example, there would be no point to run the modified program with the old header file. On the other hand, with the new header file, the program can't run until I've covered every one of the 30 critical changes scattered in the new header file so that I can't even verify the correctness of each modification until there is no more compilation error. Do you have any idea to deal with this situation?

    Read the article

  • Handling TclErrors in Python

    - by anteater7171
    In the following code I'll get the following error if I right click the window that pops up. Then go down to the very bottom entry widget then delete it's contents. It seems to be giving me a TclError. How do I go about handeling such an error? The Error Exception in Tkinter callback Traceback (most recent call last): File "C:\Python26\Lib\lib-tk\Tkinter.py", line 1410, in __call__ return self.func(*args) File "C:\Python26\CPUDEMO.py", line 503, in I TL.sclS.set(S1) File "C:\Python26\Lib\lib-tk\Tkinter.py", line 2765, in set self.tk.call(self._w, 'set', value) TclError: expected floating-point number but got "" The Code #F #PIthon.py # Import/Setup import Tkinter import psutil,time import re from PIL import Image, ImageTk from time import sleep class simpleapp_tk(Tkinter.Tk): def __init__(self,parent): Tkinter.Tk.__init__(self,parent) self.parent = parent self.initialize() def initialize(self): Widgets self.menu = Tkinter.Menu(self, tearoff = 0 ) M = [ "Options...", "Exit"] self.selectedM = Tkinter.StringVar() self.menu.add_radiobutton( label = 'Hide', variable = self.selectedM, command = self.E ) self.menu.add_radiobutton( label = 'Bump', variable = self.selectedM, command = self.E ) self.menu.add_separator() self.menu.add_radiobutton( label = 'Options...', variable = self.selectedM, command = self.E ) self.menu.add_separator() self.menu.add_radiobutton( label = 'Exit', variable = self.selectedM, command = self.E ) self.frame1 = Tkinter.Frame(self,bg='grey15',relief='ridge',borderwidth=4,width=185, height=39) self.frame1.grid() self.frame1.grid_propagate(0) self.frame1.bind( "<Button-3><ButtonRelease-3>", self.D ) self.frame1.bind( "<Button-2><ButtonRelease-2>", self.C ) self.frame1.bind( "<Double-Button-1>", self.C ) self.labelVariable = Tkinter.StringVar() self.label = Tkinter.Label(self.frame1,textvariable=self.labelVariable,fg="lightgreen",bg="grey15",borderwidth=1,font=('arial', 10, 'bold')) self.label.grid(column=1,row=0,columnspan=1,sticky='nsew') self.label.bind( "<Button-3><ButtonRelease-3>", self.D ) self.label.bind( "<Button-2><ButtonRelease-2>", self.C ) self.label.bind( "<Double-Button-1>", self.C ) self.F() self.overrideredirect(1) self.wm_attributes("-topmost", 1) global TL1 TL1 = Tkinter.Toplevel(self) TL1.wm_geometry("+0+5000") TL1.overrideredirect(1) TL1.button = Tkinter.Button(TL1,text="? CPU",fg="lightgreen",bg="grey15",activeforeground="lightgreen", activebackground='grey15',borderwidth=4,font=('Arial', 8, 'bold'),command=self.J) TL1.button.pack(ipadx=1) Events def Reset(self): self.label.configure(font=('arial', 10, 'bold'),fg='Lightgreen',bg='grey15',borderwidth=0) self.labela.configure(font=('arial', 8, 'bold'),fg='Lightgreen',bg='grey15',borderwidth=0) self.frame1.configure(bg='grey15',relief='ridge',borderwidth=4,width=224, height=50) self.label.pack(ipadx=38) def helpmenu(self): t2 = Tkinter.Toplevel(self) Tkinter.Label(t2, text='This is a help menu', anchor="w",justify="left",fg="darkgreen",bg="grey90",relief="ridge",borderwidth=5,font=('Arial', 10)).pack(fill='both', expand=1) t2.resizable(False,False) t2.title('Help') menu = Tkinter.Menu(self) t2.config(menu=menu) filemenu = Tkinter.Menu(menu) menu.add_cascade(label="| Exit |", menu=filemenu) filemenu.add_command(label="Exit", command=t2.destroy) def aboutmenu(self): t1 = Tkinter.Toplevel(self) Tkinter.Label(t1, text=' About:\n\n CPU Usage v1.0\n\n Publisher: Drew French\n Date: 05/09/10\n Email: [email protected] \n\n\n\n\n\n\n Written in Python 2.6.4', anchor="w",justify="left",fg="darkgreen",bg="grey90",relief="sunken",borderwidth=5,font=('Arial', 10)).pack(fill='both', expand=1) t1.resizable(False,False) t1.title('About') menu = Tkinter.Menu(self) t1.config(menu=menu) filemenu = Tkinter.Menu(menu) menu.add_cascade(label="| Exit |", menu=filemenu) filemenu.add_command(label="Exit", command=t1.destroy) def A (self,event): TL.entryVariable1.set(TL.sclY.get()) TL.entryVariable2.set(TL.sclX.get()) Y = TL.sclY.get() X = TL.sclX.get() self.wm_geometry("+" + str(X) + "+" + str(Y)) def B(self,event): Y1 = TL.entryVariable1.get() X1 = TL.entryVariable2.get() self.wm_geometry("+" + str(X1) + "+" + str(Y1)) TL.sclY.set(Y1) TL.sclX.set(X1) def C(self,event): s = self.wm_geometry() geomPatt = re.compile(r"(\d+)?x?(\d+)?([+-])(\d+)([+-])(\d+)") m = geomPatt.search(s) X3 = m.group(4) Y3 = m.group(6) M = int(Y3) - 150 P = M + 150 while Y3 > M: sleep(0.0009) Y3 = int(Y3) - 1 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(Y3)) sleep(2.00) while Y3 < P: sleep(0.0009) Y3 = int(Y3) + 1 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(Y3)) def D(self, event=None): self.menu.post( event.x_root, event.y_root ) def E(self): if self.selectedM.get() =='Options...': Setup global TL TL = Tkinter.Toplevel(self) menu = Tkinter.Menu(TL) TL.config(menu=menu) filemenu = Tkinter.Menu(menu) menu.add_cascade(label="| Menu |", menu=filemenu) filemenu.add_command(label="Instruction Manual...", command=self.helpmenu) filemenu.add_command(label="About...", command=self.aboutmenu) filemenu.add_separator() filemenu.add_command(label="Exit Options", command=TL.destroy) filemenu.add_command(label="Exit", command=self.destroy) helpmenu = Tkinter.Menu(menu) menu.add_cascade(label="| Help |", menu=helpmenu) helpmenu.add_command(label="Instruction Manual...", command=self.helpmenu) helpmenu.add_separator() helpmenu.add_command(label="Quick Help...", command=self.helpmenu) Title TL.label5 = Tkinter.Label(TL,text="CPU Usage: Options",anchor="center",fg="black",bg="lightgreen",relief="ridge",borderwidth=5,font=('Arial', 18, 'bold')) TL.label5.pack(padx=15,ipadx=5) X Y scale TL.separator = Tkinter.Frame(TL,height=7, bd=1, relief='ridge', bg='grey95') TL.separator.pack(pady=5,padx=5) # TL.sclX = Tkinter.Scale(TL.separator, from_=0, to=1500, orient='horizontal', resolution=1, command=self.A) TL.sclX.grid(column=1,row=0,ipadx=27, sticky='w') TL.label1 = Tkinter.Label(TL.separator,text="X",anchor="s",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label1.grid(column=0,row=0, pady=1, sticky='S') TL.sclY = Tkinter.Scale(TL.separator, from_=0, to=1500, resolution=1, command=self.A) TL.sclY.grid(column=2,row=1,rowspan=2,sticky='e', padx=4) TL.label3 = Tkinter.Label(TL.separator,text="Y",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label3.grid(column=2,row=0, padx=10, sticky='e') TL.entryVariable2 = Tkinter.StringVar() TL.entry2 = Tkinter.Entry(TL.separator,textvariable=TL.entryVariable2, fg="grey15",bg="grey90",relief="sunken",insertbackground="black",borderwidth=5,font=('Arial', 10)) TL.entry2.grid(column=1,row=1,ipadx=20, pady=10,sticky='EW') TL.entry2.bind("<Return>", self.B) TL.label2 = Tkinter.Label(TL.separator,text="X:",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label2.grid(column=0,row=1, ipadx=4, sticky='W') TL.entryVariable1 = Tkinter.StringVar() TL.entry1 = Tkinter.Entry(TL.separator,textvariable=TL.entryVariable1, fg="grey15",bg="grey90",relief="sunken",insertbackground="black",borderwidth=5,font=('Arial', 10)) TL.entry1.grid(column=1,row=2,sticky='EW') TL.entry1.bind("<Return>", self.B) TL.label4 = Tkinter.Label(TL.separator,text="Y:", anchor="center",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label4.grid(column=0,row=2, ipadx=4, sticky='W') TL.label7 = Tkinter.Label(TL.separator,text="Text Colour:",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label7.grid(column=1,row=3,stick="W",ipady=10) TL.selectedP = Tkinter.StringVar() TL.opt1 = Tkinter.OptionMenu(TL.separator, TL.selectedP,'Normal', 'White','Black', 'Blue', 'Steel Blue','Green','Light Green','Yellow','Orange' ,'Red',command=self.G) TL.opt1.config(fg="black",bg="grey90",activebackground="grey90",activeforeground="black", anchor="center",relief="raised",direction='right',font=('Arial', 10)) TL.opt1.grid(column=1,row=4,sticky='EW',padx=20,ipadx=20) TL.selectedP.set('Normal') TL.label7 = Tkinter.Label(TL.separator,text="Refresh Rate:",fg="black",bg="grey95",font=('Arial', 8 ,'bold')) TL.label7.grid(column=1,row=5,stick="W",ipady=10) TL.sclS = Tkinter.Scale(TL.separator, from_=10, to=2000, orient='horizontal', resolution=10, command=self.H) TL.sclS.grid(column=1,row=6,ipadx=27, sticky='w') TL.sclS.set(650) TL.entryVariableS = Tkinter.StringVar() TL.entryS = Tkinter.Entry(TL.separator,textvariable=TL.entryVariableS, fg="grey15",bg="grey90",relief="sunken",insertbackground="black",borderwidth=5,font=('Arial', 10)) TL.entryS.grid(column=1,row=7,ipadx=20, pady=10,sticky='EW') TL.entryS.bind("<Return>", self.I) TL.entryVariableS.set(650) # TL.resizable(False,False) TL.title('Options') geomPatt = re.compile(r"(\d+)?x?(\d+)?([+-])(\d+)([+-])(\d+)") s = self.wm_geometry() m = geomPatt.search(s) X = m.group(4) Y = m.group(6) TL.sclY.set(Y) TL.sclX.set(X) if self.selectedM.get() == 'Exit': self.destroy() if self.selectedM.get() == 'Bump': s = self.wm_geometry() geomPatt = re.compile(r"(\d+)?x?(\d+)?([+-])(\d+)([+-])(\d+)") m = geomPatt.search(s) X3 = m.group(4) Y3 = m.group(6) M = int(Y3) - 150 P = M + 150 while Y3 > M: sleep(0.0009) Y3 = int(Y3) - 1 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(Y3)) sleep(2.00) while Y3 < P: sleep(0.0009) Y3 = int(Y3) + 1 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(Y3)) if self.selectedM.get() == 'Hide': s = self.wm_geometry() geomPatt = re.compile(r"(\d+)?x?(\d+)?([+-])(\d+)([+-])(\d+)") m = geomPatt.search(s) X3 = m.group(4) Y3 = m.group(6) M = int(Y3) + 5000 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(M)) TL1.wm_geometry("+0+190") def F (self): G = round(psutil.cpu_percent(), 1) G1 = str(G) + '%' self.labelVariable.set(G1) try: S2 = TL.entryVariableS.get() except ValueError, e: S2 = 650 except NameError: S2 = 650 self.after(int(S2), self.F) def G (self,event): if TL.selectedP.get() =='Normal': self.label.config( fg = 'lightgreen' ) TL1.button.config( fg = 'lightgreen',activeforeground='lightgreen') if TL.selectedP.get() =='Red': self.label.config( fg = 'red' ) TL1.button.config( fg = 'red',activeforeground='red') if TL.selectedP.get() =='Orange': self.label.config( fg = 'orange') TL1.button.config( fg = 'orange',activeforeground='orange') if TL.selectedP.get() =='Yellow': self.label.config( fg = 'yellow') TL1.button.config( fg = 'yellow',activeforeground='yellow') if TL.selectedP.get() =='Light Green': self.label.config( fg = 'lightgreen' ) TL1.button.config( fg = 'lightgreen',activeforeground='lightgreen') if TL.selectedP.get() =='Normal': self.label.config( fg = 'lightgreen' ) TL1.button.config( fg = 'lightgreen',activeforeground='lightgreen') if TL.selectedP.get() =='Steel Blue': self.label.config( fg = 'steelblue1' ) TL1.button.config( fg = 'steelblue1',activeforeground='steelblue1') if TL.selectedP.get() =='Blue': self.label.config( fg = 'blue') TL1.button.config( fg = 'blue',activeforeground='blue') if TL.selectedP.get() =='Green': self.label.config( fg = 'darkgreen' ) TL1.button.config( fg = 'darkgreen',activeforeground='darkgreen') if TL.selectedP.get() =='White': self.label.config( fg = 'white' ) TL1.button.config( fg = 'white',activeforeground='white') if TL.selectedP.get() =='Black': self.label.config( fg = 'black') TL1.button.config( fg = 'black',activeforeground='black') def H (self,event): TL.entryVariableS.set(TL.sclS.get()) S = TL.sclS.get() def I (self,event): S1 = TL.entryVariableS.get() TL.sclS.set(S1) TL.sclS.set(TL.sclS.get()) S1 = TL.entryVariableS.get() TL.sclS.set(S1) def J (self): s = self.wm_geometry() geomPatt = re.compile(r"(\d+)?x?(\d+)?([+-])(\d+)([+-])(\d+)") m = geomPatt.search(s) X3 = m.group(4) Y3 = m.group(6) M = int(Y3) - 5000 self.update_idletasks() self.wm_geometry("+" + str(X3) + "+" + str(M)) TL1.wm_geometry("+0+5000") Loop if name == "main": app = simpleapp_tk(None) app.mainloop()

    Read the article

  • Asus P8P67 Rev. 3.1 Motherboard issues powering on and saving settings

    - by Scott
    Edit: New Information Have some updated information from the old question below: So basically my issue right now is somewhat similar, but I've been able to rule out a couple of things. I don't think this has anything to do with light on the motherboard. No matter what lights are on/off on the motherboard when the computer is off, they don't affect this issue. The main power LED on the Mobo is always lit when the power supply is turned on, and that's what matters anyway. Even when the main power LED is on, the PC will NOT boot up the first time I hit the power switch. I have to go reset the power supply (make all lights turn off on the Mobo and back on), and THEN hit the power switch. Then everything boots up. Also, the BIOS settings are reset every time this happens. Asus Tech Support told me to try jumping the power with something metal to try and rule out that it's a problem with the connectors getting power, or if it's a problem with the case power switch pins - haven't done that yet though. Any ideas? This is a lot simpler than it was before when I thought it had to do with certain LED indicators for RAM, EPU, etc. Original Question So I built my new desktop just about 3 weeks ago. I've been having a few issues which I think are all related to my motherboard, an Asus P8P67 Revision 3.1, but I'm not 100% sure as this is really the first from-scratch build I've ever done. I've posted these questions on the Asus forums, Asus Tech Support, and the Corsair forums as well as I thought it might have something to do with my power supply at one point. None of these avenues have solved my issue until now completely, so I thought I'd come here to see what you guys think. Here's what's happening: My computer is off, and I go to power it on. I press the power switch on the case (Antec Nine Hundred), and nothing seems to happen. Upon further inspection, I see that what this actually does is simply turn on the EPU LED on my motherboard, but doesn't actually boot anything up. I then have to go and flip the main power switch on the power supply off and back on. What this does is turn off all lights on the Motherboard after a few seconds, and turn them all back on (including the EPU LED that was off before I hit the power switch the first time). Now, hitting the power switch works. The machine boots up fine, and starts going through the boot up process. As a side note: My Motherboard is set to "Force BIOS", and every single time I change this to do the opposite, the next time my computer boots up that change reverts itself. I think this may be due to the fact that I am doing the hard reset on the power supply each time, but I'm not sure. I had thought that the Motherboard would keep its BIOS settings unless you did something to the Mobo itself - so this may be a related issue, or something else completely. That's basically it. Once it's on, it's on. It works fine, recognizes all of my hardware, and runs great. All fans/lights in the case work great, and I'm getting standard readings. The next time I go to shut the computer down however, I can expect the same exact process getting it up and running, including being forced to go into BIOS and exit again before I can load Windows. Another side note: If I power on my computer using the power switch DIRECTLY after shutting it down, it powers right back on (I think this is because the EPU LED light doesn't have time to turn off). It looks as if as long as the EPU LED is lit up on the motherboard before I hit the power switch on the case, the thing will boot up fine (although this doesn't explain the "Force BIOS" issue, at least it's something). Any ideas? Thanks guys. P.S. - System Specs Asus P8P67 Rev. 3.1 Motherboard Intel Core i7 2600K Processor 16GB (4x4GB) G-Skill 1600 RAM NVIDIA EVGA GTX 570 Video Card Crucial 128GB SSD HD Corsair 850W Power Supply Seagate 2TB HDD

    Read the article

  • linux raid 1: right after replacing and syncing one drive, the other disk fails - understanding what is going on with mdstat/mdadm

    - by devicerandom
    We have an old RAID 1 Linux server (Ubuntu Lucid 10.04), with four partitions. A few days ago /dev/sdb failed, and today we noticed /dev/sda had pre-failure ominous SMART signs (~4000 reallocated sector count). We replaced /dev/sdb this morning and rebuilt the RAID on the new drive, following this guide: http://www.howtoforge.com/replacing_hard_disks_in_a_raid1_array Everything went smooth until the very end. When it looked like it was finishing to synchronize the last partition, the other old one failed. At this point I am very unsure of the state of the system. Everything seems working and the files seem to be all accessible, just as if it synchronized everything, but I'm new to RAID and I'm worried about what is going on. The /proc/mdstat output is: Personalities : [raid1] [linear] [multipath] [raid0] [raid6] [raid5] [raid4] [raid10] md3 : active raid1 sdb4[2](S) sda4[0] 478713792 blocks [2/1] [U_] md2 : active raid1 sdb3[1] sda3[2](F) 244140992 blocks [2/1] [_U] md1 : active raid1 sdb2[1] sda2[2](F) 244140992 blocks [2/1] [_U] md0 : active raid1 sdb1[1] sda1[2](F) 9764800 blocks [2/1] [_U] unused devices: <none> The order of [_U] vs [U_]. Why aren't they consistent along all the array? Is the first U /dev/sda or /dev/sdb? (I tried looking on the web for this trivial information but I found no explicit indication) If I read correctly for md0, [_U] should be /dev/sda1 (down) and /dev/sdb1 (up). But if /dev/sda has failed, how can it be the opposite for md3 ? I understand /dev/sdb4 is now spare because probably it failed to synchronize it 100%, but why does it show /dev/sda4 as up? Shouldn't it be [__]? Or [_U] anyway? The /dev/sda drive now cannot even be accessed by SMART anymore apparently, so I wouldn't expect it to be up. What is wrong with my interpretation of the output? I attach also the outputs of mdadm --detail for the four partitions: /dev/md0: Version : 00.90 Creation Time : Fri Jan 21 18:43:07 2011 Raid Level : raid1 Array Size : 9764800 (9.31 GiB 10.00 GB) Used Dev Size : 9764800 (9.31 GiB 10.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 0 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:27:33 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : a3b4dbbd:859bf7f2:bde36644:fcef85e2 Events : 0.7704 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 17 1 active sync /dev/sdb1 2 8 1 - faulty spare /dev/sda1 /dev/md1: Version : 00.90 Creation Time : Fri Jan 21 18:43:15 2011 Raid Level : raid1 Array Size : 244140992 (232.83 GiB 250.00 GB) Used Dev Size : 244140992 (232.83 GiB 250.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 1 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:39:06 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : 8bcd5765:90dc93d5:cc70849c:224ced45 Events : 0.1508280 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 18 1 active sync /dev/sdb2 2 8 2 - faulty spare /dev/sda2 /dev/md2: Version : 00.90 Creation Time : Fri Jan 21 18:43:19 2011 Raid Level : raid1 Array Size : 244140992 (232.83 GiB 250.00 GB) Used Dev Size : 244140992 (232.83 GiB 250.00 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 2 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:46:44 2013 State : clean, degraded Active Devices : 1 Working Devices : 1 Failed Devices : 1 Spare Devices : 0 UUID : 2885668b:881cafed:b8275ae8:16bc7171 Events : 0.2289636 Number Major Minor RaidDevice State 0 0 0 0 removed 1 8 19 1 active sync /dev/sdb3 2 8 3 - faulty spare /dev/sda3 /dev/md3: Version : 00.90 Creation Time : Fri Jan 21 18:43:22 2011 Raid Level : raid1 Array Size : 478713792 (456.54 GiB 490.20 GB) Used Dev Size : 478713792 (456.54 GiB 490.20 GB) Raid Devices : 2 Total Devices : 2 Preferred Minor : 3 Persistence : Superblock is persistent Update Time : Tue Nov 5 17:19:20 2013 State : clean, degraded Active Devices : 1 Working Devices : 2 Failed Devices : 0 Spare Devices : 1 Number Major Minor RaidDevice State 0 8 4 0 active sync /dev/sda4 1 0 0 1 removed 2 8 20 - spare /dev/sdb4 The active sync on /dev/sda4 baffles me. I am worried because if tomorrow morning I have to replace /dev/sda, I want to be sure what should I sync with what and what is going on. I am also quite baffled by the fact /dev/sda decided to fail exactly when the raid finished resyncing. I'd like to understand what is really happening. Thanks a lot for your patience and help. Massimo

    Read the article

  • Configuring UCM cache to check for external Content Server changes

    - by Martin Deh
    Recently, I was involved in a customer scenario where they were modifying the Content Server's contributor data files directly through Content Server.  This operation of course is completely supported.  However, since the contributor data file was modified through the "backdoor", a running WebCenter Spaces page, which also used the same data file, would not get the updates immediately.  This was due to two reasons.  The first reason is that the Spaces page was using Content Presenter to display the contents of the data file. The second reason is that the Spaces application was using the "cached" version of the data file.  Fortunately, there is a way to configure cache so backdoor changes can be picked up more quickly and automatically. First a brief overview of Content Presenter.  The Content Presenter task flow enables WebCenter Spaces users with Page-Edit permissions to precisely customize the selection and presentation of content in a WebCenter Spaces application.  With Content Presenter, you can select a single item of content, contents under a folder, a list of items, or query for content, and then select a Content Presenter based template to render the content on a page in a Spaces application.  In addition to displaying the folders and the files in a Content Server, Content Presenter integrates with Oracle Site Studio to allow you to create, access, edit, and display Site Studio contributor data files (Content Server Document) in either a Site Studio region template or in a custom Content Presenter display template.  More information about creating Content Presenter Display Template can be found in the OFM Developers Guide for WebCenter Portal. The easiest way to configure the cache is to modify the WebCenter Spaces Content Server service connection setting through Enterprise Manager.  From here, under the Cache Details, there is a section to set the Cache Invalidation Interval.  Basically, this enables the cache to be monitored by the cache "sweeper" utility.  The cache sweeper queries for changes in the Content Server, and then "marks" the object in cache as "dirty".  This causes the application in turn to get a new copy of the document from the Content Server that replaces the cached version.  By default the initial value for the Cache Invalidation Interval is set to 0 (minutes).  This basically means that the sweeper is OFF.  To turn the sweeper ON, just set a value (in minutes).  The mininal value that can be set is 2 (minutes): Just a note.  In some instances, once the value of the Cache Invalidation Interval has been set (and saved) in the Enterprise Manager UI, it becomes "sticky" and the interval value cannot be set back to 0.  The good news is that this value can also be updated throught a WLST command.   The WLST command to run is as follows: setJCRContentServerConnection(appName, name, [socketType, url, serverHost, serverPort, keystoreLocation, keystorePassword, privateKeyAlias, privateKeyPassword, webContextRoot, clientSecurityPolicy, cacheInvalidationInterval, binaryCacheMaxEntrySize, adminUsername, adminPassword, extAppId, timeout, isPrimary, server, applicationVersion]) One way to get the required information for executing the command is to use the listJCRContentServerConnections('webcenter',verbose=true) command.  For example, this is the sample output from the execution: ------------------ UCM ------------------ Connection Name: UCM Connection Type: JCR External Appliction ID: Timeout: (not set) CIS Socket Type: socket CIS Server Hostname: webcenter.oracle.local CIS Server Port: 4444 CIS Keystore Location: CIS Private Key Alias: CIS Web URL: Web Server Context Root: /cs Client Security Policy: Admin User Name: sysadmin Cache Invalidation Interval: 2 Binary Cache Maximum Entry Size: 1024 The Documents primary connection is "UCM" From this information, the completed  setJCRContentServerConnection would be: setJCRContentServerConnection(appName='webcenter',name='UCM', socketType='socket', serverHost='webcenter.oracle.local', serverPort='4444', webContextRoot='/cs', cacheInvalidationInterval='0', binaryCacheMaxEntrySize='1024',adminUsername='sysadmin',isPrimary=1) Note: The Spaces managed server must be restarted for the change to take effect. More information about using WLST for WebCenter can be found here. Once the sweeper is turned ON, only cache objects that have been changed will be invalidated.  To test this out, I will go through a simple scenario.  The first thing to do is configure the Content Server so it can monitor and report on events.  Log into the Content Server console application, and under the Administration menu item, select System Audit Information.  Note: If your console is using the left menu display option, the Administration link will be located there. Under the Tracing Sections Information, add in only "system" and "requestaudit" in the Active Sections.  Check Full Verbose Tracing, check Save, then click the Update button.  Once this is done, select the View Server Output menu option.  This will change the browser view to display the log.  This is all that is needed to configure the Content Server. For example, the following is the View Server Output with the cache invalidation interval set to 2(minutes) Note the time stamp: requestaudit/6 08.30 09:52:26.001  IdcServer-68    GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.016933999955654144(secs) requestaudit/6 08.30 09:52:26.010  IdcServer-69    GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.006134999915957451(secs) requestaudit/6 08.30 09:52:26.014  IdcServer-70    GET_DOCUMENT_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.004271999932825565(secs) ... other trace info ... requestaudit/6 08.30 09:54:26.002  IdcServer-71    GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.020323999226093292(secs) requestaudit/6 08.30 09:54:26.011  IdcServer-72    GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.017928000539541245(secs) requestaudit/6 08.30 09:54:26.017  IdcServer-73    GET_DOCUMENT_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.010185999795794487(secs) Now that the tracing logs are reporting correctly, the next step is set up the Spaces app to test the sweeper. I will use 2 different pages that will use Content Presenter task flows.  Each task flow will use a different custom Content Presenter display template, and will be assign 2 different contributor data files (document that will be in the cache).  The pages at run time appear as follows: Initially, when the Space pages containing the content is loaded in the browser for the first time, you can see the tracing information in the Content Server output viewer. requestaudit/6 08.30 11:51:12.030 IdcServer-129 CLEAR_SERVER_OUTPUT [dUser=weblogic] 0.029171999543905258(secs) requestaudit/6 08.30 11:51:12.101 IdcServer-130 GET_SERVER_OUTPUT [dUser=weblogic] 0.025721000507473946(secs) requestaudit/6 08.30 11:51:26.592 IdcServer-131 VCR_GET_DOCUMENT_BY_NAME [dID=919][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][RevisionSelectionMethod=LatestReleased][IsJava=1] 0.21525299549102783(secs) requestaudit/6 08.30 11:51:27.117 IdcServer-132 VCR_GET_CONTENT_TYPES [dUser=sysadmin][IsJava=1] 0.5059549808502197(secs) requestaudit/6 08.30 11:51:27.146 IdcServer-133 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.03360399976372719(secs) requestaudit/6 08.30 11:51:27.169 IdcServer-134 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.008806000463664532(secs) requestaudit/6 08.30 11:51:27.204 IdcServer-135 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.013265999965369701(secs) requestaudit/6 08.30 11:51:27.384 IdcServer-136 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.18119299411773682(secs) requestaudit/6 08.30 11:51:27.533 IdcServer-137 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.1519480049610138(secs) requestaudit/6 08.30 11:51:27.634 IdcServer-138 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.10827399790287018(secs) requestaudit/6 08.30 11:51:27.687 IdcServer-139 VCR_GET_CONTENT_TYPE [dUser=sysadmin][IsJava=1] 0.059702999889850616(secs) requestaudit/6 08.30 11:51:28.271 IdcServer-140 GET_USER_PERMISSIONS [dUser=weblogic][IsJava=1] 0.006703000050038099(secs) requestaudit/6 08.30 11:51:28.285 IdcServer-141 GET_ENVIRONMENT [dUser=sysadmin][IsJava=1] 0.010893999598920345(secs) requestaudit/6 08.30 11:51:30.433 IdcServer-142 GET_SERVER_OUTPUT [dUser=weblogic] 0.017318999394774437(secs) requestaudit/6 08.30 11:51:41.837 IdcServer-143 VCR_GET_DOCUMENT_BY_NAME [dID=508][dDocName=113_ES][dDocTitle=Landing Home][dUser=weblogic][RevisionSelectionMethod=LatestReleased][IsJava=1] 0.15937699377536774(secs) requestaudit/6 08.30 11:51:42.781 IdcServer-144 GET_FILE [dID=326][dDocName=WEBCENTERORACL000315][dDocTitle=Duke][dUser=anonymous][RevisionSelectionMethod=LatestReleased][dSecurityGroup=Public][xCollectionID=0] 0.16288499534130096(secs) The highlighted sections show where the 2 data files DF_UCMCACHETESTER (P1 page) and 113_ES (P2 page) were called by the (Spaces) VCR connection to the Content Server. The most important line to notice is the VCR_GET_DOCUMENT_BY_NAME invocation.  On subsequent refreshes of these 2 pages, you will notice (after you refresh the Content Server's View Server Output) that there are no further traces of the same VCR_GET_DOCUMENT_BY_NAME invocations.  This is because the pages are getting the documents from the cache. The next step is to go through the "backdoor" and change one of the documents through the Content Server console.  This operation can be done by first locating the data file document, and from the Content Information page, select Edit Data File menu option.   This invokes the Site Studio Contributor, where the modifications can be made. Refreshing the Content Server View Server Output, the tracing displays the operations perform on the document.  requestaudit/6 08.30 11:56:59.972 IdcServer-255 SS_CHECKOUT_BY_NAME [dID=922][dDocName=DF_UCMCACHETESTER][dUser=weblogic][dSecurityGroup=Public] 0.05558200180530548(secs) requestaudit/6 08.30 11:57:00.065 IdcServer-256 SS_GET_CONTRIBUTOR_CONFIG [dID=922][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][dSecurityGroup=Public][xCollectionID=0] 0.08632399886846542(secs) requestaudit/6 08.30 11:57:00.470 IdcServer-259 DOC_INFO_BY_NAME [dID=922][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][dSecurityGroup=Public][xCollectionID=0] 0.02268899977207184(secs) requestaudit/6 08.30 11:57:10.177 IdcServer-264 GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.007652000058442354(secs) requestaudit/6 08.30 11:57:10.181 IdcServer-263 GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.01868399977684021(secs) requestaudit/6 08.30 11:57:10.187 IdcServer-265 GET_DOCUMENT_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.009367000311613083(secs) (internal)/6 08.30 11:57:26.118 IdcServer-266 File to be removed: /oracle/app/admin/domains/webcenter/ucm/cs/vault/~temp/703253295.xml (internal)/6 08.30 11:57:26.121 IdcServer-266 File to be removed: /oracle/app/admin/domains/webcenter/ucm/cs/vault/~temp/703253295.xml requestaudit/6 08.30 11:57:26.122 IdcServer-266 SS_SET_ELEMENT_DATA [dID=923][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][dSecurityGroup=Public][xCollectionID=0][StatusCode=0][StatusMessage=Successfully checked in content item 'DF_UCMCACHETESTER'.] 0.3765290081501007(secs) requestaudit/6 08.30 11:57:30.710 IdcServer-267 DOC_INFO_BY_NAME [dID=923][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][dSecurityGroup=Public][xCollectionID=0] 0.07942699640989304(secs) requestaudit/6 08.30 11:57:30.733 IdcServer-268 SS_GET_CONTRIBUTOR_STRINGS [dUser=weblogic] 0.0044570001773536205(secs) After a few moments and refreshing the P1 page, the updates has been applied. Note: The refresh time may very, since the Cache Invalidation Interval (set to 2 minutes) is not determined by when changes happened.  The sweeper just runs every 2 minutes. Refreshing the Content Server View Server Output, the tracing displays the important information. requestaudit/6 08.30 11:59:10.171 IdcServer-270 GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.00952600035816431(secs) requestaudit/6 08.30 11:59:10.179 IdcServer-271 GET_FOLDER_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.011118999682366848(secs) requestaudit/6 08.30 11:59:10.182 IdcServer-272 GET_DOCUMENT_HISTORY_REPORT [dUser=sysadmin][IsJava=1] 0.007447000127285719(secs) requestaudit/6 08.30 11:59:16.885 IdcServer-273 VCR_GET_DOCUMENT_BY_NAME [dID=923][dDocName=DF_UCMCACHETESTER][dDocTitle=DF_UCMCacheTester][dUser=weblogic][RevisionSelectionMethod=LatestReleased][IsJava=1] 0.0786449983716011(secs) After the specifed interval time the sweeper is invoked, which is noted by the GET_ ... calls.  Since the history has noted the change, the next call is to the VCR_GET_DOCUMENT_BY_NAME to retrieve the new version of the (modifed) data file.  Navigating back to the P2 page, and viewing the server output, there are no further VCR_GET_DOCUMENT_BY_NAME to retrieve the data file.  This simply means that this data file was just retrieved from the cache.   Upon further review of the server output, we can see that there was only 1 request for the VCR_GET_DOCUMENT_BY_NAME: requestaudit/6 08.30 12:08:00.021 Audit Request Monitor Request Audit Report over the last 120 Seconds for server webcenteroraclelocal16200****  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor -Num Requests 8 Errors 0 Reqs/sec. 0.06666944175958633 Avg. Latency (secs) 0.02762500010430813 Max Thread Count 2  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor 1 Service VCR_GET_DOCUMENT_BY_NAME Total Elapsed Time (secs) 0.09200000017881393 Num requests 1 Num errors 0 Avg. Latency (secs) 0.09200000017881393  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor 2 Service GET_PERSONALIZED_JAVASCRIPT Total Elapsed Time (secs) 0.054999999701976776 Num requests 1 Num errors 0 Avg. Latency (secs) 0.054999999701976776  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor 3 Service GET_FOLDER_HISTORY_REPORT Total Elapsed Time (secs) 0.028999999165534973 Num requests 2 Num errors 0 Avg. Latency (secs) 0.014499999582767487  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor 4 Service GET_SERVER_OUTPUT Total Elapsed Time (secs) 0.017999999225139618 Num requests 1 Num errors 0 Avg. Latency (secs) 0.017999999225139618  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor 5 Service GET_FILE Total Elapsed Time (secs) 0.013000000268220901 Num requests 1 Num errors 0 Avg. Latency (secs) 0.013000000268220901  requestaudit/6 08.30 12:08:00.021 Audit Request Monitor ****End Audit Report*****  

    Read the article

  • How Expedia Made My New Bride Cry

    - by Lance Robinson
    Tweet this? Email Expedia and ask them to give me and my new wife our honeymoon? When Expedia followed up their failure with our honeymoon trip with a complete and total lack of acknowledgement of any responsibility for the problem and endless loops of explaining the issue over and over again - I swore that they would make it right. When they brought my new bride to tears, I got an immediate and endless supply of motivation. I hope you will help me make them make it right by posting our story on Twitter, Facebook, your blog, on Expedia itself, and when talking to your friends in person about their own travel plans.   If you are considering using them now for an important trip - reconsider. Short summary: We arrived early for a flight - but Expedia had made a mistake with the data they supplied to JetBlue and Emirates, which resulted in us not being able to check in (one leg of our trip was missing)!  At the time of this post, three people (myself, my wife, and an exceptionally patient JetBlue employee named Mary) each spent hours on the phone with Expedia.  I myself spent right at 3 hours (according to iPhone records), Lauren spent an hour and a half or so, and poor Mary was probably on the phone for a good 3.5 hours.  This is after 5 hours total at the airport.  If you add up our phone time, that is nearly 8 hours of phone time over a 5 hour period with little or no help, stall tactics (?), run-around, denial, shifting of blame, and holding. Details below (times are approximate): First, my wife and I were married yesterday - June 18th, the 3 year anniversary of our first date. She is awesome. She is the nicest person I have ever known, a ton of fun, absolutely beautiful in every way. Ok enough mushy - here are the dirty details. 2:30 AM - Early Check-in Attempt - we attempted to check-in for our flight online. Some sort of technology error on website, instructed to checkin at desk. 4:30 AM - Arrive at airport. Try to check-in at kiosk, get the same error. We got to the JetBlue desk at RDU International Airport, where Mary helped us. Mary discovered that the Expedia provided itinerary does not match the Expedia provided tickets. We are informed that when that happens American, JetBlue, and others that use the same software cannot check you in for the flight because. Why? Because the itinerary was missing a leg of our flight! Basically we were not shown in the system as definitely being able to make it home. Mary called Expedia and was put on hold by their automated system. 4:55 AM - Mary, myself, and my brand new bride all waited for about 25 minutes when finally I decided I would make a call myself on my iPhone while Mary was on the airport phone. In their automated system, I chose "make a new reservation", thinking they might answer a little more quickly than "customer service". Not surprisingly I was connected to an Expedia person within 1 minute. They informed me that they would have to forward me to a customer service specialist. I explained to them that we were already on hold for that and had been for nearly half an hour, that we were going on our honeymoon and that our flight would be leaving soon - could they please help us. "Yes, I will help you". I hand the phone to JetBlue Mary who explains the situation 3 or 4 times. Obviously I couldn't hear both ends of the conversation at this point, but the Expedia person explained what the problem was by stating exactly what Mary had just spent 15 minutes explaining. Mary calmly confirms that this is the problem, and asks Expedia to re-issue the itinerary. Expedia tells Mary that they'll have to transfer her to customer service. Mary asks for someone specific so that we get an answer this time, and goes on hold. Mary get's connected, explains the situation, and then Mary's connection gets terminated. 5:10 AM - Mary calls back to the Expedia automated system again, and we wait for about 5 minutes on hold this time before I pick up my iPhone and call Expedia again myself. Again I go to sales, a person picks up the phone in less than a minute. I explain the situation and let them know that we are now very close to missing our flight for our honeymoon, could they please help us. "Yes, I will help you". Again I give the phone to Mary who provides them with a call back number in case we get disconnected again and explains the situation again. More back and forth with Expedia doing nothing but repeating the same questions, Mary answering the questions with the same information she provided in the original explanation, and Expedia simply restating the problem. Mary again asks them to re-issue the itinerary, and explains that doing so will fix the problem. Expedia again repeats the problem instead of fixing it, and Mary's connection gets terminated. 5:20 AM - Mary again calls back to Expedia. My beautiful bride also calls on her own phone. At this point she is struggling to hold back her tears, stumbling through an explanation of all that has happened and that we are about to miss our flight. Please help us. "Yes, I will help". My beautiful bride's connection gets terminated. Ok, maybe this disconnection isn't an accident. We've now been disconnected 3 times on two different phones. 5:45 AM - I walk away and pleadingly beg a person to help me. They "escalate" the issue to "Rosy" (sp?) at Expedia. I go through the whole song and dance again with Rosy, who gives me the same treatment Mary was given. Rosy blames JetBlue for now having the correct data. Meanwhile Mary is on the phone with Emirates Air (the airline for the second leg of our trip), who agrees with JetBlue that Expedia's data isn't up to date. We are informed by two airport employees that issues like this with Expedia are not uncommon, and that the fix is simple. On the phone iwth Rosy, I ask her to re-issue the itinerary because we are about to miss our flight. She again explains the problem to me. At this point, I am standing at the window, pleading with Rosy to help us get to our honeymoon, watching our airplane. Then our airplane leaves without us. 6:03 AM - At this point we have missed our flight. Re-issuing the itinerary is no longer a solution. I ask Rosy to start from the beginning and work us up a new trip. She says that she cannot do that. She says that she needs to talk to JetBlue and Emirates and find out why we cannot check-in for our flight. I remind Rosy that our flight has already left - I just watched it taxi away - it no longer matters why (not to mention the fact that we already knew why, and have known why since 4:30 AM), and have known the solution since 4:30 AM. Rosy, can you please book a new trip? Yes, but it will cost $400. Excuse me? Now you can, but it will cost ME to fix your mistake? Rosy says that she can escalate the situation to her supervisor but that will take 1.5 hours. 6:15 AM - I told Rosy that if they had re-issued the itinerary as JetBlue asked (at 4:30 AM), my new wife and I might be on the airplane now instead of dealing with this on the phone and missing the beginning (and how much more?) of our honeymoon. Rosy said that it was not necessary to re-issue the itinerary. Out of curiosity, i asked Rosy if there was some financial burden on them to re-issue the itinerary. "No", said Rosy. I asked her if it was a large time burden on Expedia to re-issue the itinerary. "No", said Rosy. I directly asked Rosy: Why wouldn't Expedia have re-issued the itinerary when JetBlue asked? No answer. I asked Rosy: If you had re-issued the itinerary at 4:30, isn't it possible that I would be on that flight right now? She actually surprised me by answering "Yes" to that question. So I pointed out that it followed that Expedia was responsible for the fact that we missed out flight, and she immediately went into more about how the problem was with JetBlue - but now it was ALSO an Emirates Air problem as well. I tell Rosy to go ahead and escalate the issue again, and please call me back in that 1.5 hours (which how is about 1 hour and 10 minutes away). 6:30 AM - I start tweeting my frustration with iPhone. It's now pretty much impossible for us to make it to The Maldives by 3pm, which is the time at which we would need to arrive in order to be allowed service to the actual island where we are staying. Expedia has now given me the run-around for 2 hours, caused me to miss my flight, and worst of all caused my amazing new wife Lauren to miss our honeymoon. You think I was mad? No. Furious. Its ok to make mistakes - but to refuse to fix them and to ruin our honeymoon? No, not ok, Expedia. I swore right then that Expedia would make this right. 7:45 AM - JetBlue mary is still talking her tail off to other people in JetBlue and Emirates Air. Mary works it out so that if Expedia simply books a new trip, JetBlue and Emirates will both waive all the fees. Now we just have to convince Expedia to fix their mistake and get us on our way! Around this time Expedia Rosy calls me back! I inform her of the excellent work of JetBlue Mary - that JetBlue and Emirates both will waive the fees so Expedia can fix their mistake and get us going on our way. She says that she sees documentation of this in her system and that she needs to put me on hold "for 1 to 10 minutes" to talk to Emirates Air (why I'm not exactly sure). I say ok. 8:45 AM - After an hour on hold, Rosy comes on the line and asks me to hold more. I ask her to call me back. 9:35 AM - I put down the iPhone Twitter app and picks up the laptop. You think I made some noise with my iPhone? Heh 11:25 AM - Expedia follows me and sends a canned "We're sorry, DM us the details".  If you look at their Twitter feed, 16 out of the most recent 20 tweets are exactly the same canned response.  The other 4?  Ads.  Um - #MultiFAIL? To Expedia:  You now have had (as explained above) 8 hours of 3 different people explaining our situation, you know the email address of our Expedia account, you know my web blog, you know my Twitter address, you know my phone number.  You also know how upset you have made both me and my new bride by treating us with such a ... non caring, scripted, uncooperative, argumentative, and possibly even deceitful manner.  In the wise words of the great Kenan Thompson of SNL: "FIX IT!".  And no, I'm NOT going away until you make this right. Period. 11:45 AM - Expedia corporate office called.  The woman I spoke to was very nice and apologetic.  She listened to me tell the story again, she says she understands the problem and she is going to work to resolve it.  I don't have any details on what exactly that resolution might me, she said she will call me back in 20 minutes.  She found out about the problem via Twitter.  Thank you Twitter, and all of you who helped.  Hopefully social media will win my wife and I our honeymoon, and hopefully Expedia will encourage their customer service teams treat their customers properly. 12:22 PM - Spoke to Fran again from Expedia corporate office.  She has a flight for us tonight.  She is booking it now.  We will arrive at our honeymoon destination of beautiful Veligandu Island Resort only 1 day late.  She cannot confirm today, but she expects that Expedia will pay for the lost honeymoon night.  Thank you everyone for your help.  I will reflect more on this whole situation and confirm its resolution after our flight is 100% confirmed.  For now, I'm going to take a breather and go kiss my wonderful wife! 1:50 PM - Have not yet received the promised phone call.  We did receive an email with a new itinerary for a flight but the booking is not for specific seats, so there is no guarantee that my wife and I will be able to sit together.  With the original booking I carefully selected our seats for every segment of our trip.  I decided to call into the phone number that Fran from the Expedia corporate office gave me.  Its automated voice system identified itself as "Tier 3 Support".  I am currently still on hold with them, I have not gotten through to a human yet. 1:55 PM - Fran from Expedia called me back.  She confirmed us as booked.  She called the airlines to confirm.  Unfortunately, Expedia was unwilling or unable to allow us any type of seat selection.  It is possible that i won't get to sit next to the woman I married less than a day ago on our 40 total hours of flight time (there and back).  In addition, our seats could be the worst seats on the planes, with no reclining seat back or right next to the restroom.  Despite this fact (which in my opinion is huge), the horrible inconvenience, the hours at the airport, and the negative Internet publicity that Expedia is receiving, Expedia declined to offer us any kind of upgrade or to mark us as SFU (suitable for upgrade).  Since they didn't offer - I asked, and was rejected.  I am grateful to finally be heading in the right direction, but not only did Expedia horribly botch this job from the very beginning, they followed that botch job with near zero customer service, followed by a verbally apologetic but otherwise half-hearted resolution.  If this works out favorably for us, great.  If not - I'm not done making noise, Expedia.  You owe us, and I expect you to make it right.  You haven't quite done that yet. Thanks - Thank you to Twitter.  Thanks to all those who sympathize with us and helped us get the attention of Expedia, since three people (one of them an airline employee) using Expedia's normal channels of communication for many hours didn't help.  Thanks especially to my PowerShell and Sharepoint friends, my local friends, and those connectors who encouraged me and spread my story. 5:15 PM - Love Wins - After all this, Lauren and I are exhausted.  We both took a short nap, and when we woke up we talked about the last 24 hours.  It was a big, amazing, story-filled 24 hours.  I said that Expedia won, but Lauren said no.  She pointed out how lucky we are.  We are in love and married.  We have wonderful family and friends.  We are both hard-working successful people who love what they do.  We get to go to an amazing exotic destination for our honeymoon like Veligandu in The Maldives...  That's a lot of good.  Expedia didn't win.  This was (is) a big loss for Expedia.  It is a public blemish for all to see.  But Lauren and I did win, big time.  Expedia may not have made things right - but things are right for us.  Post in progress... I will relay any further comments (or lack of) from Expedia soon, as well as an update on confirmation of their repayment of our lost resort room rates.  I'll also post a picture of us on our honeymoon as soon as I can!

    Read the article

  • Building The Right SharePoint Team For Your Organization

    - by Mark Rackley
    I see the question posted fairly often asking what kind SharePoint team an organization should have. How many people do I need? What roles do I need to fill? What is best for my organization? Well, just like every other answer in SharePoint, the correct answer is “it depends”. Do you ever get sick of hearing that??? I know I do… So, let me give you my thoughts and opinions based upon my experience and what I’ve seen and let you come to your own conclusions. What are the possible SharePoint roles? I guess the first thing you need to understand are the different roles that exist in SharePoint (and their are LOTS). Remember, SharePoint is a massive beast and you will NOT find one person who can do it all. If you are hoping to find that person you will be sorely disappointed. For the most part this is true in SharePoint 2007 and 2010. However, generally things are improved in 2010 and easier for junior individuals to grasp. SharePoint Administrator The absolutely positively only role that you should not be without no matter the size of your organization or SharePoint deployment is a SharePoint administrator. These guys are essential to keeping things running and figuring out what’s wrong when things aren’t running well. These unsung heroes do more before 10 am than I do all day. The bad thing is, when these guys are awesome, you don’t even know they exist because everything is running so smoothly. You should definitely invest some time and money here to make sure you have some competent if not rockstar help. You need an admin who truly loves SharePoint and will go that extra mile when necessary. Let me give you a real world example of what I’m talking about: We have a rockstar admin… and I’m sure she’s sick of my throwing her name around so she’ll just have to live with remaining anonymous in this post… sorry Lori… Anyway! A couple of weeks ago our Server teams came to us and said Hi Lori, I’m finalizing the MOSS servers and doing updates that require a restart; can I restart them? Seems like a harmless request from your server team does it not? Sure, go ahead and apply the patches and reboot during our scheduled maintenance window. No problem? right? Sounded fair to me… but no…. not to our fearless SharePoint admin… I need a complete list of patches that will be applied. There is an update that is out there that will break SharePoint… KB973917 is the patch that has been shown to cause issues. What? You mean Microsoft released a patch that would actually adversely affect SharePoint? If we did NOT have a rockstar admin, our server team would have applied these patches and then when some problem occurred in SharePoint we’d have to go through the fun task of tracking down exactly what caused the issue and resolve it. How much time would that have taken? If you have a junior SharePoint admin or an admin who’s not out there staying on top of what’s going on you could have spent days tracking down something so simple as applying a patch you should not have applied. I will even go as far to say the only SharePoint rockstar you NEED in your organization is a SharePoint admin. You can always outsource really complicated development projects or bring in a rockstar contractor every now and then to make sure you aren’t way off track in other areas. For your day-to-day sanity and to keep SharePoint running smoothly, you need an awesome Admin. Some rockstars in this category are: Ben Curry, Mike Watson, Joel Oleson, Todd Klindt, Shane Young, John Ferringer, Sean McDonough, and of course Lori Gowin. SharePoint Developer Another essential role for your SharePoint deployment is a SharePoint developer. Things do start to get a little hazy here and there are many flavors of “developers”. Are you writing custom code? using SharePoint Designer? What about SharePoint Branding?  Are all of these considered developers? I would say yes. Are they interchangeable? I’d say no. Development in SharePoint is such a large beast in itself. I would say that it’s not so large that you can’t know it all well, but it is so large that there are many people who specialize in one particular category. If you are lucky enough to have someone on staff who knows it all well, you better make sure they are well taken care of because those guys are ready-made to move over to a consulting role and charge you 3 times what you are probably paying them. :) Some of the all-around rockstars are Eric Shupps, Andrew Connell (go Razorbacks), Rob Foster, Paul Schaeflein, and Todd Bleeker SharePoint Power User/No-Code Solutions Developer These SharePoint Swiss Army Knives are essential for quick wins in your organization. These people can twist the out-of-the-box functionality to make it do things you would not even imagine. Give these guys SharePoint Designer, jQuery, InfoPath, and a little time and they will create views, dashboards, and KPI’s that will blow your mind away and give your execs the “wow” they are looking for. Not only can they deliver that wow factor, but they can mashup, merge, and really help make your SharePoint application usable and deliver an overall better user experience. Before you hand off a project to your SharePoint Custom Code developer, let one of these rockstars look at it and show you what they can do (in probably less time). I would say the second most important role you can fill in your organization is one of these guys. Rockstars in this category are Christina Wheeler, Laura Rogers, Jennifer Mason, and Mark Miller SharePoint Developer – Custom Code If you want to really integrate SharePoint into your legacy systems, or really twist it and make it bend to your will, you are going to have to open up Visual Studio and write some custom code.  Remember, SharePoint is essentially just a big, huge, ginormous .NET application, so you CAN write code to make it do ANYTHING, but do you really want to spend the time and effort to do so? At some point with every other form of SharePoint development you are going to run into SOME limitation (SPD Workflows is the big one that comes to mind). If you truly want to knock down all the walls then custom development is the way to go. PLEASE keep in mind when you are looking for a custom code developer that a .NET developer does NOT equal a SharePoint developer. Just SOME of the things these guys write are: Custom Workflows Custom Web Parts Web Service functionality Import data from legacy systems Export data to legacy systems Custom Actions Event Receivers Service Applications (2010) These guys are also the ones generally responsible for packaging everything up into solution packages (you are doing that, right?). Rockstars in this category are Phil Wicklund, Christina Wheeler, Geoff Varosky, and Brian Jackett. SharePoint Branding “But it LOOKS like SharePoint!” Somebody call the WAAAAAAAAAAAAHMbulance…   Themes, Master Pages, Page Layouts, Zones, and over 2000 styles in CSS.. these guys not only have to be comfortable with all of SharePoint’s quirks and pain points when branding, but they have to know it TWICE for publishing and non-publishing sites.  Not only that, but these guys really need to have an eye for graphic design and be able to translate the ramblings of business into something visually stunning. They also have to be comfortable with XSLT, XML, and be able to hand off what they do to your custom developers for them to package as solutions (which you are doing, right?). These rockstars include Heater Waterman, Cathy Dew, and Marcy Kellar SharePoint Architect SharePoint Architects are generally SharePoint Admins or Developers who have moved into more of a BA role? Is that fair to say? These guys really have a grasp and understanding for what SharePoint IS and what it can do. These guys help you structure your farms to meet your needs and help you design your applications the correct way. It’s always a good idea to bring in a rockstar SharePoint Architect to do a sanity check and make sure you aren’t doing anything stupid.  Most organizations probably do not have a rockstar architect on staff. These guys are generally brought in at the deployment of a farm, upgrade of a farm, or for large development projects. I personally also find architects very useful for sitting down with the business to translate their needs into what SharePoint can do. A good architect will be able to pick out what can be done out-of-the-box and what has to be custom built and hand those requirements to the development Staff. Architects can generally fill in as an admin or a developer when needed. Some rockstar architects are Rick Taylor, Dan Usher, Bill English, Spence Harbar, Neil Hodgkins, Eric Harlan, and Bjørn Furuknap. Other Roles / Specialties On top of all these other roles you also get these people who specialize in things like Reporting, BDC (BCS in 2010), Search, Performance, Security, Project Management, etc... etc... etc... Again, most organizations will not have one of these gurus on staff, they’ll just pay out the nose for them when they need them. :) SharePoint End User Everyone else in your organization that touches SharePoint falls into this category. What they actually DO in SharePoint is determined by your governance and what permissions you give these guys. Hopefully you have these guys on a fairly short leash and are NOT giving them access to tools like SharePoint Designer. Sadly end users are the ones who truly make your deployment a success by using it, but are also your biggest enemy in breaking it.  :)  We love you guys… really!!! Okay, all that’s fine and dandy, but what should MY SharePoint team look like? It depends! Okay… Are you just doing out of the box team sites with no custom development? Then you are probably fine with a great Admin team and a great No-Code Solution Development team. How many people do you need? Depends on how busy you can keep them. Sorry, can’t answer the question about numbers without knowing your specific needs. I can just tell you who you MIGHT need and what they will do for you. I’ll leave you with what my ideal SharePoint Team would look like for a particular scenario: Farm / Organization Structure Dev, QA, and 2 Production Farms. 5000 – 10000 Users Custom Development and Integration with legacy systems Team Sites, My Sites, Intranet, Document libraries and overall company collaboration Team Rockstar SharePoint Administrator 2-3 junior SharePoint Administrators SharePoint Architect / Lead Developer 2 Power User / No-Code Solution Developers 2-3 Custom Code developers Branding expert With a team of that size and skill set, they should be able to keep a substantial SharePoint deployment running smoothly and meet your business needs. This does NOT mean that you would not need to bring in contract help from time to time when you need an uber specialist in one area. Also, this team assumes there will be ongoing development for the life of your SharePoint farm. If you are just going to be doing sporadic custom development, it might make sense to partner with an awesome firm that specializes in that sort of work (I can give you the name of a couple if you are interested).  Again though, the size of your team depends on the number of requests you are receiving and how much active deployment you are doing. So, don’t bring in a team that looks like this and then yell at me because they are sitting around with nothing to do or are so overwhelmed that nothing is getting done. I do URGE you to take the proper time to asses your needs and determine what team is BEST for your organization. Also, PLEASE PLEASE PLEASE do not skimp on the talent. When it comes to SharePoint you really do get what you pay for when it comes to employees, contractors, and software.  SharePoint can become absolutely critical to your business and because you skimped on hiring a developer he created a web part that brings down the farm because he doesn’t know what he’s doing, or you hire an admin who thinks it’s fine to stick everything in the same Content Database and then can’t figure out why people are complaining. SharePoint can be an enormous blessing to an organization or it’s biggest curse. Spend the time and money to do it right, or be prepared to spending even more time and money later to fix it.

    Read the article

  • Beware Sneaky Reads with Unique Indexes

    - by Paul White NZ
    A few days ago, Sandra Mueller (twitter | blog) asked a question using twitter’s #sqlhelp hash tag: “Might SQL Server retrieve (out-of-row) LOB data from a table, even if the column isn’t referenced in the query?” Leaving aside trivial cases (like selecting a computed column that does reference the LOB data), one might be tempted to say that no, SQL Server does not read data you haven’t asked for.  In general, that’s quite correct; however there are cases where SQL Server might sneakily retrieve a LOB column… Example Table Here’s a T-SQL script to create that table and populate it with 1,000 rows: CREATE TABLE dbo.LOBtest ( pk INTEGER IDENTITY NOT NULL, some_value INTEGER NULL, lob_data VARCHAR(MAX) NULL, another_column CHAR(5) NULL, CONSTRAINT [PK dbo.LOBtest pk] PRIMARY KEY CLUSTERED (pk ASC) ); GO DECLARE @Data VARCHAR(MAX); SET @Data = REPLICATE(CONVERT(VARCHAR(MAX), 'x'), 65540);   WITH Numbers (n) AS ( SELECT ROW_NUMBER() OVER (ORDER BY (SELECT 0)) FROM master.sys.columns C1, master.sys.columns C2 ) INSERT LOBtest WITH (TABLOCKX) ( some_value, lob_data ) SELECT TOP (1000) N.n, @Data FROM Numbers N WHERE N.n <= 1000; Test 1: A Simple Update Let’s run a query to subtract one from every value in the some_value column: UPDATE dbo.LOBtest WITH (TABLOCKX) SET some_value = some_value - 1; As you might expect, modifying this integer column in 1,000 rows doesn’t take very long, or use many resources.  The STATITICS IO and TIME output shows a total of 9 logical reads, and 25ms elapsed time.  The query plan is also very simple: Looking at the Clustered Index Scan, we can see that SQL Server only retrieves the pk and some_value columns during the scan: The pk column is needed by the Clustered Index Update operator to uniquely identify the row that is being changed.  The some_value column is used by the Compute Scalar to calculate the new value.  (In case you are wondering what the Top operator is for, it is used to enforce SET ROWCOUNT). Test 2: Simple Update with an Index Now let’s create a nonclustered index keyed on the some_value column, with lob_data as an included column: CREATE NONCLUSTERED INDEX [IX dbo.LOBtest some_value (lob_data)] ON dbo.LOBtest (some_value) INCLUDE ( lob_data ) WITH ( FILLFACTOR = 100, MAXDOP = 1, SORT_IN_TEMPDB = ON ); This is not a useful index for our simple update query; imagine that someone else created it for a different purpose.  Let’s run our update query again: UPDATE dbo.LOBtest WITH (TABLOCKX) SET some_value = some_value - 1; We find that it now requires 4,014 logical reads and the elapsed query time has increased to around 100ms.  The extra logical reads (4 per row) are an expected consequence of maintaining the nonclustered index. The query plan is very similar to before (click to enlarge): The Clustered Index Update operator picks up the extra work of maintaining the nonclustered index. The new Compute Scalar operators detect whether the value in the some_value column has actually been changed by the update.  SQL Server may be able to skip maintaining the nonclustered index if the value hasn’t changed (see my previous post on non-updating updates for details).  Our simple query does change the value of some_data in every row, so this optimization doesn’t add any value in this specific case. The output list of columns from the Clustered Index Scan hasn’t changed from the one shown previously: SQL Server still just reads the pk and some_data columns.  Cool. Overall then, adding the nonclustered index hasn’t had any startling effects, and the LOB column data still isn’t being read from the table.  Let’s see what happens if we make the nonclustered index unique. Test 3: Simple Update with a Unique Index Here’s the script to create a new unique index, and drop the old one: CREATE UNIQUE NONCLUSTERED INDEX [UQ dbo.LOBtest some_value (lob_data)] ON dbo.LOBtest (some_value) INCLUDE ( lob_data ) WITH ( FILLFACTOR = 100, MAXDOP = 1, SORT_IN_TEMPDB = ON ); GO DROP INDEX [IX dbo.LOBtest some_value (lob_data)] ON dbo.LOBtest; Remember that SQL Server only enforces uniqueness on index keys (the some_data column).  The lob_data column is simply stored at the leaf-level of the non-clustered index.  With that in mind, we might expect this change to make very little difference.  Let’s see: UPDATE dbo.LOBtest WITH (TABLOCKX) SET some_value = some_value - 1; Whoa!  Now look at the elapsed time and logical reads: Scan count 1, logical reads 2016, physical reads 0, read-ahead reads 0, lob logical reads 36015, lob physical reads 0, lob read-ahead reads 15992.   CPU time = 172 ms, elapsed time = 16172 ms. Even with all the data and index pages in memory, the query took over 16 seconds to update just 1,000 rows, performing over 52,000 LOB logical reads (nearly 16,000 of those using read-ahead). Why on earth is SQL Server reading LOB data in a query that only updates a single integer column? The Query Plan The query plan for test 3 looks a bit more complex than before: In fact, the bottom level is exactly the same as we saw with the non-unique index.  The top level has heaps of new stuff though, which I’ll come to in a moment. You might be expecting to find that the Clustered Index Scan is now reading the lob_data column (for some reason).  After all, we need to explain where all the LOB logical reads are coming from.  Sadly, when we look at the properties of the Clustered Index Scan, we see exactly the same as before: SQL Server is still only reading the pk and some_value columns – so what’s doing the LOB reads? Updates that Sneakily Read Data We have to go as far as the Clustered Index Update operator before we see LOB data in the output list: [Expr1020] is a bit flag added by an earlier Compute Scalar.  It is set true if the some_value column has not been changed (part of the non-updating updates optimization I mentioned earlier). The Clustered Index Update operator adds two new columns: the lob_data column, and some_value_OLD.  The some_value_OLD column, as the name suggests, is the pre-update value of the some_value column.  At this point, the clustered index has already been updated with the new value, but we haven’t touched the nonclustered index yet. An interesting observation here is that the Clustered Index Update operator can read a column into the data flow as part of its update operation.  SQL Server could have read the LOB data as part of the initial Clustered Index Scan, but that would mean carrying the data through all the operations that occur prior to the Clustered Index Update.  The server knows it will have to go back to the clustered index row to update it, so it delays reading the LOB data until then.  Sneaky! Why the LOB Data Is Needed This is all very interesting (I hope), but why is SQL Server reading the LOB data?  For that matter, why does it need to pass the pre-update value of the some_value column out of the Clustered Index Update? The answer relates to the top row of the query plan for test 3.  I’ll reproduce it here for convenience: Notice that this is a wide (per-index) update plan.  SQL Server used a narrow (per-row) update plan in test 2, where the Clustered Index Update took care of maintaining the nonclustered index too.  I’ll talk more about this difference shortly. The Split/Sort/Collapse combination is an optimization, which aims to make per-index update plans more efficient.  It does this by breaking each update into a delete/insert pair, reordering the operations, removing any redundant operations, and finally applying the net effect of all the changes to the nonclustered index. Imagine we had a unique index which currently holds three rows with the values 1, 2, and 3.  If we run a query that adds 1 to each row value, we would end up with values 2, 3, and 4.  The net effect of all the changes is the same as if we simply deleted the value 1, and added a new value 4. By applying net changes, SQL Server can also avoid false unique-key violations.  If we tried to immediately update the value 1 to a 2, it would conflict with the existing value 2 (which would soon be updated to 3 of course) and the query would fail.  You might argue that SQL Server could avoid the uniqueness violation by starting with the highest value (3) and working down.  That’s fine, but it’s not possible to generalize this logic to work with every possible update query. SQL Server has to use a wide update plan if it sees any risk of false uniqueness violations.  It’s worth noting that the logic SQL Server uses to detect whether these violations are possible has definite limits.  As a result, you will often receive a wide update plan, even when you can see that no violations are possible. Another benefit of this optimization is that it includes a sort on the index key as part of its work.  Processing the index changes in index key order promotes sequential I/O against the nonclustered index. A side-effect of all this is that the net changes might include one or more inserts.  In order to insert a new row in the index, SQL Server obviously needs all the columns – the key column and the included LOB column.  This is the reason SQL Server reads the LOB data as part of the Clustered Index Update. In addition, the some_value_OLD column is required by the Split operator (it turns updates into delete/insert pairs).  In order to generate the correct index key delete operation, it needs the old key value. The irony is that in this case the Split/Sort/Collapse optimization is anything but.  Reading all that LOB data is extremely expensive, so it is sad that the current version of SQL Server has no way to avoid it. Finally, for completeness, I should mention that the Filter operator is there to filter out the non-updating updates. Beating the Set-Based Update with a Cursor One situation where SQL Server can see that false unique-key violations aren’t possible is where it can guarantee that only one row is being updated.  Armed with this knowledge, we can write a cursor (or the WHILE-loop equivalent) that updates one row at a time, and so avoids reading the LOB data: SET NOCOUNT ON; SET STATISTICS XML, IO, TIME OFF;   DECLARE @PK INTEGER, @StartTime DATETIME; SET @StartTime = GETUTCDATE();   DECLARE curUpdate CURSOR LOCAL FORWARD_ONLY KEYSET SCROLL_LOCKS FOR SELECT L.pk FROM LOBtest L ORDER BY L.pk ASC;   OPEN curUpdate;   WHILE (1 = 1) BEGIN FETCH NEXT FROM curUpdate INTO @PK;   IF @@FETCH_STATUS = -1 BREAK; IF @@FETCH_STATUS = -2 CONTINUE;   UPDATE dbo.LOBtest SET some_value = some_value - 1 WHERE CURRENT OF curUpdate; END;   CLOSE curUpdate; DEALLOCATE curUpdate;   SELECT DATEDIFF(MILLISECOND, @StartTime, GETUTCDATE()); That completes the update in 1280 milliseconds (remember test 3 took over 16 seconds!) I used the WHERE CURRENT OF syntax there and a KEYSET cursor, just for the fun of it.  One could just as well use a WHERE clause that specified the primary key value instead. Clustered Indexes A clustered index is the ultimate index with included columns: all non-key columns are included columns in a clustered index.  Let’s re-create the test table and data with an updatable primary key, and without any non-clustered indexes: IF OBJECT_ID(N'dbo.LOBtest', N'U') IS NOT NULL DROP TABLE dbo.LOBtest; GO CREATE TABLE dbo.LOBtest ( pk INTEGER NOT NULL, some_value INTEGER NULL, lob_data VARCHAR(MAX) NULL, another_column CHAR(5) NULL, CONSTRAINT [PK dbo.LOBtest pk] PRIMARY KEY CLUSTERED (pk ASC) ); GO DECLARE @Data VARCHAR(MAX); SET @Data = REPLICATE(CONVERT(VARCHAR(MAX), 'x'), 65540);   WITH Numbers (n) AS ( SELECT ROW_NUMBER() OVER (ORDER BY (SELECT 0)) FROM master.sys.columns C1, master.sys.columns C2 ) INSERT LOBtest WITH (TABLOCKX) ( pk, some_value, lob_data ) SELECT TOP (1000) N.n, N.n, @Data FROM Numbers N WHERE N.n <= 1000; Now here’s a query to modify the cluster keys: UPDATE dbo.LOBtest SET pk = pk + 1; The query plan is: As you can see, the Split/Sort/Collapse optimization is present, and we also gain an Eager Table Spool, for Halloween protection.  In addition, SQL Server now has no choice but to read the LOB data in the Clustered Index Scan: The performance is not great, as you might expect (even though there is no non-clustered index to maintain): Table 'LOBtest'. Scan count 1, logical reads 2011, physical reads 0, read-ahead reads 0, lob logical reads 36015, lob physical reads 0, lob read-ahead reads 15992.   Table 'Worktable'. Scan count 1, logical reads 2040, physical reads 0, read-ahead reads 0, lob logical reads 34000, lob physical reads 0, lob read-ahead reads 8000.   SQL Server Execution Times: CPU time = 483 ms, elapsed time = 17884 ms. Notice how the LOB data is read twice: once from the Clustered Index Scan, and again from the work table in tempdb used by the Eager Spool. If you try the same test with a non-unique clustered index (rather than a primary key), you’ll get a much more efficient plan that just passes the cluster key (including uniqueifier) around (no LOB data or other non-key columns): A unique non-clustered index (on a heap) works well too: Both those queries complete in a few tens of milliseconds, with no LOB reads, and just a few thousand logical reads.  (In fact the heap is rather more efficient). There are lots more fun combinations to try that I don’t have space for here. Final Thoughts The behaviour shown in this post is not limited to LOB data by any means.  If the conditions are met, any unique index that has included columns can produce similar behaviour – something to bear in mind when adding large INCLUDE columns to achieve covering queries, perhaps. Paul White Email: [email protected] Twitter: @PaulWhiteNZ

    Read the article

< Previous Page | 281 282 283 284 285 286 287 288 289 290 291 292  | Next Page >