Search Results

Search found 32254 results on 1291 pages for 'model view presenter'.

Page 503/1291 | < Previous Page | 499 500 501 502 503 504 505 506 507 508 509 510  | Next Page >

  • Database commands

    - by user12609425
    Ops Center has two database options - you can have Ops Center automatically install a database on the Enterprise Controller system, or you can use your own database on any system you choose. If you use your own database, it's obviously important to make sure that this database is running smoothly. You have a few tools that can help you do this. The first is the ecadm command. This command has a variety of subcommands that let you view and control the status of the Enterprise Controller. Two subcommands in particular are relevant to the database: ecadm verify-db: This subcommand verifies that the database is reachable and that the schemas are configured with the proper permissions. Use the -v option if you want more details; the command is normally terse if the DB is configured correctly. ecadm sqlplus -r: This subcommand opens an sqlplus console connection to the database. The -r option makes this console read-only, which isn't necessary, but is generally a good idea. You can also view the database contents using Oracle SQL Developer or other tools. The Accessing Core Product Data how-to describes this process.

    Read the article

  • How can I make Excel documents open in different windows?

    - by Eugene
    Office 2007, Windows Server 2008 x64. How can I make Excel so that when I double-click a document, it opens in a new Excel instance, so that I can easily view them side-by-side as separate windows and not using the View-Arrange All functionality? Now I have to go to the task bar, click on one document to see it and then click on the other document in the task bar to switch to that one. As the alternative, I close one document, open a new Excel window, then drag the document in there. Thank you.

    Read the article

  • How do I instal Skype on iMac G5 Ubuntu 12.04.1?

    - by Sergiu
    So I have this imac G5 model 8.7 running powerpc64. I can't install any of the .debs they have on skype.com because they're the wrong architecture and I also tried adding the partner repo for the software center but when I do sudo apt-get install skype it says it can't find the package as if the repo I just added doesn't exist. I added the repo with the following command: sudo add-apt-repository "deb http://archive.canonical.com/ $(lsb_release -sc) partner" I'm running Ubuntu Release 12.04 (precise) 32-bit Kernel Linux 3.2.0-31-powerpc64-smp GNOME 3.4.2

    Read the article

  • Ubuntu 12.04 - NVIDIA GT 525M Error

    - by talha06
    I removed all packages associated with NVIDIA from Synaptic Package Manager and installed nvidia-current. When I click NVIDIA-Settings, gives an error like this: You do not appear to be using the NVIDIA X driver. Please edit your X configuration file (just run `nvidia-xconfig` as root), and restart the X server. Please help me, I need your helps. My laptop model is Dell Inspiron N5110 using Nvidia GeForce GT525M..

    Read the article

  • Any addin or tool to open multiple web sites on one page?

    - by David.Chu.ca
    Is there any addin or tools available for web browser to view multiple web sites on one page or tab? Or is there any web service to allow me to customize one page for multiple web sites? For example, within a page, there are some tools to allow me to define regions like a table or several frames. Then I'll be able to add url to each cell or frames. After settings, I'll be able to view several sites on one tab or page. Not sure if there is any tools available?

    Read the article

  • Build 2013&ndash;Keynote Thoughts

    - by D'Arcy Lussier
    Originally posted on: http://geekswithblogs.net/dlussier/archive/2013/06/26/153243.aspxSome thoughts on the Build 2013 keynote. They Listened to Feedback while Keeping to their Plans I am one of the people in the “bring back the start menu” camp. I want my start menu. I *like* my start menu. Microsoft heard that and put it back, fantastic. But they implemented it in a way that still pushes the Windows 8 UI – and I’m actually pretty happy with it. When you hit the Start menu, you get the live-tiles displayed overlaying the desktop. But you can also swipe from the bottom to get the “all-applications” view. This, in essense, is really what those that like the Start Menu want. I believe it was mentioned that you can configure the all-applications view to be the default. They’re Committed to Improving Windows 8 The commitment to rapid deployments Ballmer talked about is crucial to Windows 8’s success. They need to keep it evolving quickly to maintain the interest of users and developers. I think the little improvements they showed are excellent (hands-free mode, multi window docking, better multi-monitor support, new developer controls, etc.). Hardware Vendors are Committed to Windows 8 They showed off a number of new hardware products (Windows 8 and Windows Phone). The Surface’s introduction to the market has done nothing to dissuade their hardware partners. Bing as a Platform is Huge for Developers!!! This was the biggest take-away from the keynote! What the team is doing with Bing not as a search engine but as a developer API is very impressive! I’m going to be diving into this over the rest of Build so watch more blog posts coming on it. Azure, Office 365, and other topics will be covered at tomorrow’s keynote. So far, great kick off to Build. Now on to sessions! D

    Read the article

  • Minecraft mob spawning coding?

    - by Richard
    I recently discovered how to change the game (with MCP) and now I'd like to do my first "big" change to the game, creating new mobs. I already made their skin, the model, the AI and added a new entityID to the mob list. I just need to know how to make them spawn normally under similar conditions to zombies and skeletons. Thanks in advance! :D EDIT: Also, if anyone knows it, post tutorials about minecraft code editing, that would be great.

    Read the article

  • JavaOne Content Available for Free

    - by Tori Wieldt
    JavaOne content is available in video in three sizes, depending on if you want to have a sip, have a drink, or go to the proverbial firehose. Tall (Keynote Highlights) Go to the JavaOne playlist on the YouTube Java channel for highlights of the JavaOne Keynotes.  Grande (Keynotes in Full) Go to the Oracle Media Network JavaOne 2012 channel to view the keynotes in full (Community Keynote coming soon). Venti (All Sessions, BOFs and Tutorials) To view slides paired with audio of each session, go to the JavaOne content catalog (JavaOne homepage, click on JavaOne Technical Sessions) and select a session. If a video is available, you'll see "Media" in the right column. Look under "Presentation Download" to get the slides. Sessions are being made available as quickly as possible. "It's exciting to see Oracle take community stewardship so seriously," said Sharat Chander, Group Director for Java Technology Outreach. "Making all JavaOne sessions on video available online for free will helps make the future Java for everyone." Thanks to Oracle for funding this and providing to it to the Java Community for free.

    Read the article

  • Tool or website or process to display previews of website templates residing in archive files?

    - by Tony_Henrich
    I have hundreds of website templates in rar or zip files. To view any of them I have to extract the archive to a temporary folder and then view the template in there. It's a time consuming manual process to do this for each template Is there a tool which enables me to quickly preview the templates in the files? OR (if I extract each template into a separate folder off a master folder) A web app which can enable previewing of each template by automatically creating a link or a preview image (similar to template sites) of the home page for? OR any method to preview the templates in the fastest convenient way possible?

    Read the article

  • Include page url in php error_log

    - by scatteredbomb
    I'm working on sorting out all my php errors. I've just setup so that all errors go into one log so I can view them. The problem is that I can't really see what page someone trying to access that gave them an error. Of course I can see the file where the error was generated, but these are often includes used on other pages - so I can't really see what URL a person was attempting to view when this error was displayed, so I can't really replicate the problem. Any suggestions? Is it possible to include the actual page URL?

    Read the article

  • What kind of permission is this? (Groups+Roles)

    - by Jorge
    I'm starting to need an access control for roles in my app. I don't know much of this, but I understand how vBulletin works: I create groups, then give permissions to groups. I think that what I need is the Role Bases Access Control (RBAC) , but i'm not sure, because I need groups to give permissions instead of single users (Maybe it's not that complicated to achieve). Example of what I'm thinking: Given a post: Editor's Group has permission to view it before it's published. Editor's Group has permission to edit its content. Public Group (Default) has not permission to view it before it's published. Admin Group has permission to delete the post. So basically I wan't orientation about if RBAC is what I need. And also, how would it be good to store group membership in a user, for example, would be good to have: ID NAME PASSWORD GROUPS (1, MyName, MyPassword, 1/2/3/4/5) and explode it via PHP or one registry for every Group membership in a table named permissions, example: USERID, USERGROUP values (1, 1), (1, 2) Maybe should be the second way because of the formal norms but I didn't study yet Databases 1 at college.

    Read the article

  • RPi and Java Embedded GPIO: It all begins with hardware

    - by hinkmond
    So, you want to connect low-level peripherals (like blinky-blinky LEDs) to your Raspberry Pi and use Java Embedded technology to program it, do you? You sick foolish masochist. No, just kidding! That's awesome! You've come to the right place. I'll step you though it. And, as with many embedded projects, it all begins with hardware. So, the first thing to do is to get acquainted with the GPIO header on your RPi board. A "header" just means a thingy with a bunch of pins sticking up from it where you can connect wires. See the the red box outline in the photo. Now, there are many ways to connect to that header outlined by the red box in the photo (which the RPi folks call the P1 header). One way is to use a breakout kit like the one at Adafruit. But, we'll just use jumper wires in this example. So, to connect jumper wires to the header you need a map of where to connect which wire. That's why you need to study the pinout in the photo. That's your map for connecting wires. But, as with many things in life, it's not all that simple. RPi folks have made things a little tricky. There are two revisions of the P1 header pinout. One for older boards (RPi boards made before Sep 2012), which is called Revision 1. And, one for those fancy 512MB boards that were shipped after Sep 2012, which is called Revision 2. So, first make sure which board you have: either you have the Model A or B with 128MB or 256MB built before Sep 2012 and you need to look at the pinout for Rev. 1, or you have the Model B with 512MB and need to look at Rev. 2. That's all you need for now. More to come... Hinkmond

    Read the article

  • Sharepoint 2010 can't find domain users when granting permissions

    - by quani
    I'm trying to grant permissions to other people to view a SharePoint site but when granting permissions it uses "Check Names" and claims any user or group that is part of a domain does not exist. It does this if I try granting permissions to the team site or in central admin BUT if I try to add someone to Farm Administrators in Central admin then all of the sudden it can find all domain users. Why is it finding domain users in that one context but not others? It is supposed to be using NTLM authentication and has Windows configured as the authentication provider (And IIS is configured to use NTLM). What's even more strange is I enabled Anonymous Access for the team site which I thought would allow anyone to view it but others say they can't access it.

    Read the article

  • Different Ways to Restore a SQL Server Database

    This article describes the SQL Server database restore principles for full backups, differential backups and transaction log backups and how to perform the restores to get to a particular point in time. This tip describes SQL Server database restore principles on a database that is using the FULL recovery model. FREE eBook – "45 Database Performance Tips for Developers"Improve your database performance with 45 tips from SQL Server MVPs and industry experts. Get the eBook here.

    Read the article

  • Munin help... configure graphs to show different time periods

    - by PiZzL3
    I'm running a server with Munin installed through WHM cPanel. I've been googling around and I can't seem to find out how to do the following: A: Change a graph to show a different time period (currently I can only view, day, week, month, year). I would like to possibly view per minute, per hour... or specific intervals such as per 30 minutes, per 4 hours, etc... B(optional, but preferred): Add new graphs with the above criteria(A) above. I am a novice with all things that will possibly be required to do this (such as ssh). I wish I could just go edit a php file as I'm highly experienced in that subject. Thanks for any help!

    Read the article

  • Imperative vs. component based programming [closed]

    - by AlexW
    I've been thinking about how programming and more specifically the teaching of programming is advocated amongst the community (online). Often I've heard that Ruby and RoR is an ideal platform for learning to program. I completely disagree... RoR and Ruby are based on the application of the component based paradigm, which means they are ideal for rapid application development. This is much like the MVC model in PHP and ASP.NET But, learning a proper imperative language like Java or C/C++ (or even Perl and PHP) is the only way for a new programmer to explore logic itself, and not get too bogged down in architectural concerns like the need for separation of concerns, and the preference for components. Maybe it's a personal preference thing. I rather think that the most interesting aspects to programming are the procedural bits of code I write that actually do stuff rather than the project planning, and modelling that comes about from fully object oriented engineering or simply using the MVC model. I know this may sound confused to some of you. I feel strongly though that the best way for programming to be taught is through imperative and procedural methods. Architectural (component) methods come later, if at all. After all, none of the amazing algorithms that exist were based on OOP practice! It's all procedural code when it comes to the 'magic'. OOP is useful in creating products and utilities. Algorithms are what makes things happen, and move data around, and so imperative (and/or procedural) code are what matters most. When I see programmers recommending Ruby on Rails to newbie developers, I think it's just so wrong. Just because you write less code with Ruby does not make it easier to do! It's the opposite... you have to know loads more to appreciate its succinct nature. New coders who really want to understand the nuts and bolts of coding need to go away and figure out writing methods/functions (i.e. imperative programming) and working in procedural style, in order to grasp the fundamentals, first, before looking into architectural ways of working. So, my question is: should Ruby ever be recommended as a first language? I think no (obviously)... what arguments are there for it?

    Read the article

  • OpenGL, objects disappear at a certain z distance

    - by smoth190
    I'm writing a managed OpenGL library in C++, and I'm having a pretty annoying problem. Whenever I set an objects position to -2.0 or lower, the object disappears. And at distances 0 through -1.9, it doesn't appear to move away from them camera. I have a world matrix (which is multiplied by the objects position to move it), a view matrix (which is just the identity matrix currently) and a projection matrix, which is setup like this: FOV: 45.0f Aspect Ratio: 1 zNear: 0.1f zFar: 100.0f using GLMs glm::perspective method. The order the matrices are multiplied by in the shader is world, view, projection, then position. I can move the object along the X and Y axis perfectly fine. I have depth testing enabled, using GL_LEQUAL. I can change the actually vertices positions to anything I want, and they move away from the camera or towards it perfectly fine. It just seems to be the world matrix acting up. I'm using glm::mat4 for the world matrix, and glm::vec3 for positions. Whats going on here? I'm also using OpenGL 3.1, GLSL version 140 (1.4?).

    Read the article

  • Can my machine run Ubuntu( kubuntu | xubuntu | mint ) 12.04 WELL?

    - by Steve
    I have a 9 year old computer packing the hardware listed below. My question is, can I run 12.04 ( Ubuntu, Kubuntu, MINT or Xubuntu ) WELL? I was running Ubuntu 10.10 and I upgraded to 12.04 by going through each release via the update manager: 11.04 - 11.10 -12.04 During the installation process for 12.04 I saw an error message that there was an error installing and setting up part of the kernel. Later, when I tried installing a package in synaptic, I got another error message mentioning the kernel. When I rebooted, I got told somehting about my video and graphics was not configured properly and that I would have to do it manually ( like I know ). It gave me the option to enter the system in low graphics mode, but it just hanged. I had an old livecd of Xubuntu 10.10 around so I used that to get into my computer and copy data over to an external hard drive. I think tried to install Xubuntu 10.10 from the livecd, with the option "download updates" checked. The install process moved along a bit, then halted for about 5 hours. I rebooted my machine and tried the Xubuntu 10.10 installer WITHOUT the option to "download updates". The install completed in about 15 minutes. So, all of that is making me wonder if there is someting about 12.04 that does not like my hardware. I'm willing to try again, but only if I know I will not have to spend hours just to get to an error message and a hosed up system like I did last night. I also think I have a lot more RAM than is being reported in the output below. I had extra ram installed last year. I'm not good with the command line readouts, but there seems like there should be a lot more. I wasn't thrilled with Unity. I am willing to try Kubuntu 12.04. Will I run into the same problems? What is the highest version of a *ubuntu can I upgrade to? Thanks CPU Model: Intel(R) Pentium(R) 4 CPU 2.53GHz Frequency: 2533.223 MHz L2 Cache: 512 KB Bogomips: 5066.44 Numbering: family(15) model(2) stepping(7) Flags: fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe up pebs bts cid RAM ~$ free -mt total used free shared buffers cached Mem: 1506 891 615 0 91 521 -/+ buffers/cache: 278 1227 Swap: 1609 0 1609 Total: 3116 891 2225 Video Card 01:00.0 VGA compatible controller: nVidia Corporation NV18 [GeForce4 MX 440 AGP 8x] (rev a2) (prog-if 00 [VGA controller]) Subsystem: ASUSTeK Computer Inc. V9180 Magic Flags: bus master, 66MHz, medium devsel, latency 32, IRQ 16 Memory at fd000000 (32-bit, non-prefetchable) [size=16M] Memory at f0000000 (32-bit, prefetchable) [size=64M] Expansion ROM at fe9e0000 [disabled] [size=128K] Kernel driver in use: nouveau Kernel modules: nouveau, nvidiafb Motherboard Intel 845PE ATX 533FSB DDR333 USB2

    Read the article

  • Quaternions, Axis Angles and Rotation Matrices. Which of these should I use for FP Camera?

    - by Afonso Lage
    After 2 weeks of reading many math formulas and such I know what is a Quaternion, an Axis Angles and Matrices. I have made my own math libary (Java) to use on my game (LWJGL). But I'm really confused about all this. I want to have a 3D first person camera. The move (translation) is working fine but the rotation isnt working like I need. I need a camera to rotate arround world Axis and not about its own axis. But even using Quaternions, this doesnt work and no matter how much I read about Euler Angles, everybody says to me dont touch on it! This is a little piece of code that i'm using to make the rotation: Quaternion qPitch = Quaternion.createFromAxis(cameraRotate.x, 1.0f, 0.0f, 0.0f); Quaternion qYaw = Quaternion.createFromAxis(cameraRotate.y, 0.0f, 1.0f, 0.0f); this.multiplicate(qPitch.toMatrix4f().toArray()); this.multiplicate(qYaw.toMatrix4f().toArray()); Where this is a Matrix4f view matrix and cameraRotate is a Vector3f that just handle the angles to rotate obtained from mouse move. So I think I'm doing everything right: Translate the view Matrix Rotate the Move Matrix So, after reading all this, I just want to know: To obtain a correct first person camera rotate, I must need to use Quaternios to make the rotations, but how to rotate around world axis? Thanks for reading it. Best regards, Afonso Lage

    Read the article

  • [EF + Oracle] Inserting Data (Sequences) (2/2)

    - by JTorrecilla
    Prologue In the previous chapter we have see how to create DB records with EF, now we are going to Some Questions about Oracle.   ORACLE One characteristic from SQL Server that differs from Oracle is “Identity”. To all that has not worked with SQL Server, this property, that applies to Integer Columns, lets indicate that there is auto increment columns, by that way it will be filled automatically, without writing it on the insert statement. In EF with SQL Server, the properties whose match with Identity columns, will be filled after invoking SaveChanges method. In Oracle DB, there is no Identity Property, but there is something similar. Sequences Sequences are DB objects, that allow to create auto increment, but there are not related directly to a Table. The syntax is as follows: name, min value, max value and begin value. 1: CREATE SEQUENCE nombre_secuencia 2: INCREMENT BY numero_incremento 3: START WITH numero_por_el_que_empezara 4: MAXVALUE valor_maximo | NOMAXVALUE 5: MINVALUE valor_minimo | NOMINVALUE 6: CYCLE | NOCYCLE 7: ORDER | NOORDER 8:    How to get sequence value? To obtain the next val from the sequence: 1: SELECT nb_secuencia.Nextval 2: From Dual Due to there is no direct way to indicate that a column is related to a sequence, there is several ways to imitate the behavior: Use a Trigger (DB), Use Stored Procedures or Functions(…) or my particularly option. EF model, only, imports Table Objects, Stored Procedures or Functions, but not sequences. By that, I decide to create my own extension Method to invoke Next Val from a sequence: 1: public static class EFSequence 2: { 3: public static int GetNextValue(this ObjectContext contexto, string SequenceName) 4: { 5: string Connection = ConfigurationManager.ConnectionStrings["JTorrecillaEntities2"].ConnectionString; 6: Connection=Connection.Substring(Connection.IndexOf(@"connection string=")+19); 7: Connection = Connection.Remove(Connection.Length - 1, 1); 8: using (IDbConnection con = new Oracle.DataAccess.Client.OracleConnection(Connection)) 9: { 10: using (IDbCommand cmd = con.CreateCommand()) 11: { 12: con.Open(); 13: cmd.CommandText = String.Format("Select {0}.nextval from DUAL", SequenceName); 14: return Convert.ToInt32(cmd.ExecuteScalar()); 15: } 16: } 17:  18: } 19: } This Object Context’s extension method are going to invoke a query with the Sequence indicated by parameter. It takes the connection strings from the App settings, removing the meta data, that was created by VS when generated the EF model. And then, returns the next value from the Sequence. The next value of a Sequence is unique, by that, when some concurrent users are going to create records in the DB using the sequence will not get duplicates. This is my own implementation, I know that it could be several ways to do and better ways. If I find any other way, I promise to post it. To use the example is needed to add a reference to the Oracle (ODP.NET) dll.

    Read the article

  • Unable to connect to sharedfax error

    - by Owl
    I'm running SBS2k on a machine with a 56k modem for all outgoing faxes. We have a terminal server running server 2003 enterprise and in printers the sharedfax shows "unable to connect status". When you right-click, properties, it gives you an error saying the spool service is not running and you cannot view properties. The spooler service is indeed running on the 2k3 server because I can view the other printer's properties..and on 2k I show the Microsoft Shared Fax service AND Print Spooler as started. Here's the tricky part, I can add the shared fax printer on a domain computer locally (Windows XP) and it will show ready status and I can send faxes, they show up on the 2k server's fax console manager as in progress. I've sent a couple from a few different workstations today successfully, but still unable to send from 2k3 terminal services.

    Read the article

  • Role based access to resources for a RESTful service

    - by mutex
    I'm still wrapping my head around REST, but I wonder if someone can help with any suggestions or approaches to role based access control for a RESTful service, particularly from the point of view of securing the data and how the URLs might look. It's probably best to consider an example: Say I have a REST service for Customers, and want to split the users of this REST service into Admin, Editor and Reader roles: Admins can change all attributes of a Customer resource Editors can change only some Readers can only view them. Access control rights are assigned to the Customers entities individually. So for example a user of the service might have admin rights to Customers 1,2 and 3 but Editor access to 4,5 and Reader access to 7,8,9. Now consider the user calling the service. What is a good way to seperate the list of Customers for the current User? GET /Customer - this might get a list of all customers that the current user has Admin\Editor\Reader access to. But then on each Customer the consumer would need an indication of what role they have. Or would it be "better" having something like GET /Customer/Admin - return all customers the current user has Admin access to. Just looking for some high level pointers or reading on a decent way to secure\filter the resources based on roles of the current user.

    Read the article

  • Why C++ people loves multithreading when it comes to performances?

    - by user1849534
    I have a question, it's about why programmers seems to love concurrency and multi-threaded programs in general. I'm considering 2 main approach here: an async approach basically based on signals, or just an async approach as called by many papers and languages like the new C# 5.0 for example, and a "companion thread" that maanges the policy of your pipeline a concurrent approach or multi-threading approach I will just say that I'm thinking about the hardware here and the worst case scenario, and I have tested this 2 paradigms myself, the async paradigm is a winner at the point that I don't get why people 90% of the time talk about concurrency when they wont to speed up things or make a good use of their resources. I have tested multi-threaded programs and async program on an old machine with an Intel quad-core that doesn't offer a memory controller inside the CPU, the memory is managed entirely by the motherboard, well in this case performances are horrible with a multi-threaded application, even a relatively low number of threads like 3-4-5 can be a problem, the application is unresponsive and is just slow and unpleasant. A good async approach is, on the other hand, probably not faster but it's not worst either, my application just waits for the result and doesn't hangs, it's responsive and there is a much better scaling going on. I have also discovered that a context change in the threading world it's not that cheap in real world scenario, it's infact quite expensive especially when you have more than 2 threads that need to cycle and swap among each other to be computed. On modern CPUs the situation it's not really that different, the memory controller it's integrated but my point is that an x86 CPUs is basically a serial machine and the memory controller works the same way as with the old machine with an external memory controller on the motherboard. The context switch is still a relevant cost in my application and the fact that the memory controller it's integrated or that the newer CPU have more than 2 core it's not bargain for me. For what i have experienced the concurrent approach is good in theory but not that good in practice, with the memory model imposed by the hardware, it's hard to make a good use of this paradigm, also it introduces a lot of issues ranging from the use of my data structures to the join of multiple threads. Also both paradigms do not offer any security abut when the task or the job will be done in a certain point in time, making them really similar from a functional point of view. According to the X86 memory model, why the majority of people suggest to use concurrency with C++ and not just an async aproach ? Also why not considering the worst case scenario of a computer where the context switch is probably more expensive than the computation itself ?

    Read the article

  • Why is my concurrency capacity so low for my web app on a LAMP EC2 instance?

    - by AMF
    I come from a web developer background and have been humming along building my PHP app, using the CakePHP framework. The problem arose when I began the ab (Apache Bench) testing on the Amazon EC2 instance in which the app resides. I'm getting pretty horrendous average page load times, even though I'm running a c1.medium instance (2 cores, 2GB RAM), and I think I'm doing everything right. I would run: ab -n 200 -c 20 http://localhost/heavy-but-view-cached-page.php Here are the results: Concurrency Level: 20 Time taken for tests: 48.197 seconds Complete requests: 200 Failed requests: 0 Write errors: 0 Total transferred: 392111200 bytes HTML transferred: 392047600 bytes Requests per second: 4.15 [#/sec] (mean) Time per request: 4819.723 [ms] (mean) Time per request: 240.986 [ms] (mean, across all concurrent requests) Transfer rate: 7944.88 [Kbytes/sec] received While the ab test is running, I run VMStat, which shows that Swap stays at 0, CPU is constantly at 80-100% (although I'm not sure I can trust this on a VM), RAM utilization ramps up to about 1.6G (leaving 400M free). Load goes up to about 8 and site slows to a crawl. Here's what I think I'm doing right on the code side: In Chrome browser uncached pages typically load in 800-1000ms, and cached pages load in 300-500ms. Not stunning, but not terrible either. Thanks to view caching, there might be at most one DB query per page-load to write session data. So we can rule out a DB bottleneck. I have APC on. I am using Memcached to serve the view cache and other site caches. xhprof code profiler shows that cached pages take up 10MB-40MB in memory and 100ms - 1000ms in wall time. Pages that would be the worst offenders would look something like this in xhprof: Total Incl. Wall Time (microsec): 330,143 microsecs Total Incl. CPU (microsecs): 320,019 microsecs Total Incl. MemUse (bytes): 36,786,192 bytes Total Incl. PeakMemUse (bytes): 46,667,008 bytes Number of Function Calls: 5,195 My Apache config: KeepAlive On MaxKeepAliveRequests 100 KeepAliveTimeout 3 <IfModule mpm_prefork_module> StartServers 5 MinSpareServers 5 MaxSpareServers 10 MaxClients 120 MaxRequestsPerChild 1000 </IfModule> Is there something wrong with the server? Some gotcha with the EC2? Or is it my code? Some obvious setting I should look into? Too many DNS lookups? What am I missing? I really want to get to 1,000 concurrency capacity, but at this rate, it ain't gonna happen.

    Read the article

  • Expected time for lazy evaluation with nested functions?

    - by Matt_JD
    A colleague and I are doing a free R course, although I believe this is a more general lazy evaluation issue, and have found a scenario that we have discussed briefly and I'd like to find out the answer from a wider community. The scenario is as follows (pseudo code): wrapper => function(thing) { print => function() { write(thing) } } v = createThing(1, 2, 3) w = wrapper(v) v = createThing(4, 5, 6) w.print() // Will print 4, 5, 6 thing. v = create(7, 8, 9) w.print() // Will print 4, 5, 6 because "thing" has now been evaluated. Another similar situation is as follows: // Using the same function as above v = createThing(1, 2, 3) v = wrapper(v) w.print() // The wrapper function incestuously includes itself. Now I understand why this happens but where my colleague and I differ is on what should happen. My colleague's view is that this is a bug and the evaluation of the passed in argument should be forced at the point it is passed in so that the returned "w" function is fixed. My view is that I would prefer his option myself, but that I realise that the situation we are encountering is down to lazy evaluation and this is just how it works and is more a quirk than a bug. I am not actually sure of what would be expected, hence the reason I am asking this question. I think that function comments could express what will happen, or leave it to be very lazy, and if the coder using the function wants the argument evaluated then they can force it before passing it in. So, when working with lazy evaulation, what is the practice for the time to evaluate an argument passed, and stored, inside a function?

    Read the article

< Previous Page | 499 500 501 502 503 504 505 506 507 508 509 510  | Next Page >