Search Results

Search found 17045 results on 682 pages for 'high cpu usage'.

Page 39/682 | < Previous Page | 35 36 37 38 39 40 41 42 43 44 45 46  | Next Page >

  • SQL SERVER – How to Ignore Columnstore Index Usage in Query

    - by pinaldave
    Earlier I wrote about SQL SERVER – Fundamentals of Columnstore Index and very first question I received in email was as following. “We are using SQL Server 2012 CTP3 and so far so good. In our data warehouse solution we have created 1 non-clustered columnstore index on our large fact table. We have very unique situation but your article did not cover it. We are running few queries on our fact table which is working very efficiently but there is one query which earlier was running very fine but after creating this non-clustered columnstore index this query is running very slow. We dropped the columnstore index and suddenly this one query is running fast but other queries which were benefited by this columnstore index it is running slow. Any workaround in this situation?” In summary the question in simple words “How can we ignore using columnstore index in selective queries?” Very interesting question – you can use I can understand there may be the cases when columnstore index is not ideal and needs to be ignored the same. You can use the query hint IGNORE_NONCLUSTERED_COLUMNSTORE_INDEX to ignore the columnstore index. SQL Server Engine will use any other index which is best after ignoring the columnstore index. Here is the quick script to prove the same. We will first create sample database and then create columnstore index on the same. Once columnstore index is created we will write simple query. This query will use columnstore index. We will then show the usage of the query hint. USE AdventureWorks GO -- Create New Table CREATE TABLE [dbo].[MySalesOrderDetail]( [SalesOrderID] [int] NOT NULL, [SalesOrderDetailID] [int] NOT NULL, [CarrierTrackingNumber] [nvarchar](25) NULL, [OrderQty] [smallint] NOT NULL, [ProductID] [int] NOT NULL, [SpecialOfferID] [int] NOT NULL, [UnitPrice] [money] NOT NULL, [UnitPriceDiscount] [money] NOT NULL, [LineTotal] [numeric](38, 6) NOT NULL, [rowguid] [uniqueidentifier] NOT NULL, [ModifiedDate] [datetime] NOT NULL ) ON [PRIMARY] GO -- Create clustered index CREATE CLUSTERED INDEX [CL_MySalesOrderDetail] ON [dbo].[MySalesOrderDetail] ( [SalesOrderDetailID]) GO -- Create Sample Data Table -- WARNING: This Query may run upto 2-10 minutes based on your systems resources INSERT INTO [dbo].[MySalesOrderDetail] SELECT S1.* FROM Sales.SalesOrderDetail S1 GO 100 -- Create ColumnStore Index CREATE NONCLUSTERED COLUMNSTORE INDEX [IX_MySalesOrderDetail_ColumnStore] ON [MySalesOrderDetail] (UnitPrice, OrderQty, ProductID) GO Now we have created columnstore index so if we run following query it will use for sure the same index. -- Select Table with regular Index SELECT ProductID, SUM(UnitPrice) SumUnitPrice, AVG(UnitPrice) AvgUnitPrice, SUM(OrderQty) SumOrderQty, AVG(OrderQty) AvgOrderQty FROM [dbo].[MySalesOrderDetail] GROUP BY ProductID ORDER BY ProductID GO We can specify Query Hint IGNORE_NONCLUSTERED_COLUMNSTORE_INDEX as described in following query and it will not use columnstore index. -- Select Table with regular Index SELECT ProductID, SUM(UnitPrice) SumUnitPrice, AVG(UnitPrice) AvgUnitPrice, SUM(OrderQty) SumOrderQty, AVG(OrderQty) AvgOrderQty FROM [dbo].[MySalesOrderDetail] GROUP BY ProductID ORDER BY ProductID OPTION (IGNORE_NONCLUSTERED_COLUMNSTORE_INDEX) GO Let us clean up the database. -- Cleanup DROP INDEX [IX_MySalesOrderDetail_ColumnStore] ON [dbo].[MySalesOrderDetail] GO TRUNCATE TABLE dbo.MySalesOrderDetail GO DROP TABLE dbo.MySalesOrderDetail GO Again, make sure that you use hint sparingly and understanding the proper implication of the same. Make sure that you test it with and without hint and select the best option after review of your administrator. Here is the question for you – have you started to use SQL Server 2012 for your validation and development (not on production)? It will be interesting to know the answer. Reference: Pinal Dave (http://blog.SQLAuthority.com) Filed under: PostADay, SQL, SQL Authority, SQL Index, SQL Optimization, SQL Performance, SQL Query, SQL Server, SQL Tips and Tricks, T SQL, Technology

    Read the article

  • Skewed: a rotating camera in a simple CPU-based voxel raycaster/raytracer

    - by voxelizr
    TL;DR -- in my first simple software voxel raycaster, I cannot get camera rotations to work, seemingly correct matrices notwithstanding. The result is skewed: like a flat rendering, correctly rotated, however distorted and without depth. (While axis-aligned ie. unrotated, depth and parallax are as expected.) I'm trying to write a simple voxel raycaster as a learning exercise. This is purely CPU based for now until I figure out how things work exactly -- fow now, OpenGL is just (ab)used to blit the generated bitmap to the screen as often as possible. Now I have gotten to the point where a perspective-projection camera can move through the world and I can render (mostly, minus some artifacts that need investigation) perspective-correct 3-dimensional views of the "world", which is basically empty but contains a voxel cube of the Stanford Bunny. So I have a camera that I can move up and down, strafe left and right and "walk forward/backward" -- all axis-aligned so far, no camera rotations. Herein lies my problem. Screenshot #1: correct depth when the camera is still strictly axis-aligned, ie. un-rotated. Now I have for a few days been trying to get rotation to work. The basic logic and theory behind matrices and 3D rotations, in theory, is very clear to me. Yet I have only ever achieved a "2.5 rendering" when the camera rotates... fish-eyey, bit like in Google Streetview: even though I have a volumetric world representation, it seems --no matter what I try-- like I would first create a rendering from the "front view", then rotate that flat rendering according to camera rotation. Needless to say, I'm by now aware that rotating rays is not particularly necessary and error-prone. Still, in my most recent setup, with the most simplified raycast ray-position-and-direction algorithm possible, my rotation still produces the same fish-eyey flat-render-rotated style looks: Screenshot #2: camera "rotated to the right by 39 degrees" -- note how the blue-shaded left-hand side of the cube from screen #2 is not visible in this rotation, yet by now "it really should"! Now of course I'm aware of this: in a simple axis-aligned-no-rotation-setup like I had in the beginning, the ray simply traverses in small steps the positive z-direction, diverging to the left or right and top or bottom only depending on pixel position and projection matrix. As I "rotate the camera to the right or left" -- ie I rotate it around the Y-axis -- those very steps should be simply transformed by the proper rotation matrix, right? So for forward-traversal the Z-step gets a bit smaller the more the cam rotates, offset by an "increase" in the X-step. Yet for the pixel-position-based horizontal+vertical-divergence, increasing fractions of the x-step need to be "added" to the z-step. Somehow, none of my many matrices that I experimented with, nor my experiments with matrix-less hardcoded verbose sin/cos calculations really get this part right. Here's my basic per-ray pre-traversal algorithm -- syntax in Go, but take it as pseudocode: fx and fy: pixel positions x and y rayPos: vec3 for the ray starting position in world-space (calculated as below) rayDir: vec3 for the xyz-steps to be added to rayPos in each step during ray traversal rayStep: a temporary vec3 camPos: vec3 for the camera position in world space camRad: vec3 for camera rotation in radians pmat: typical perspective projection matrix The algorithm / pseudocode: // 1: rayPos is for now "this pixel, as a vector on the view plane in 3d, at The Origin" rayPos.X, rayPos.Y, rayPos.Z = ((fx / width) - 0.5), ((fy / height) - 0.5), 0 // 2: rotate around Y axis depending on cam rotation. No prob since view plane still at Origin 0,0,0 rayPos.MultMat(num.NewDmat4RotationY(camRad.Y)) // 3: a temp vec3. planeDist is -0.15 or some such -- fov-based dist of view plane from eye and also the non-normalized, "in axis-aligned world" traversal step size "forward into the screen" rayStep.X, rayStep.Y, rayStep.Z = 0, 0, planeDist // 4: rotate this too -- 0,zstep should become some meaningful xzstep,xzstep rayStep.MultMat(num.NewDmat4RotationY(CamRad.Y)) // set up direction vector from still-origin-based-ray-position-off-rotated-view-plane plus rotated-zstep-vector rayDir.X, rayDir.Y, rayDir.Z = -rayPos.X - me.rayStep.X, -rayPos.Y, rayPos.Z + rayStep.Z // perspective projection rayDir.Normalize() rayDir.MultMat(pmat) // before traversal, the ray starting position has to be transformed from origin-relative to campos-relative rayPos.Add(camPos) I'm skipping the traversal and sampling parts -- as per screens #1 through #3, those are "basically mostly correct" (though not pretty) -- when axis-aligned / unrotated.

    Read the article

  • Clever memory usage through the years

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

    Read the article

  • Simple OpenGL program major slow down at high resolution

    - by Grieverheart
    I have created a small OpenGL 3.3 (Core) program using freeglut. The whole geometry is two boxes and one plane with some textures. I can move around like in an FPS and that's it. The problem is I face a big slow down of fps when I make my window large (i.e. above 1920x1080). I have monitors GPU usage when in full-screen and it shows GPU load of nearly 100% and Memory Controller load of ~85%. When at 600x600, these numbers are at about 45%, my CPU is also at full load. I use deferred rendering at the moment but even when forward rendering, the slow down was nearly as severe. I can't imagine my GPU is not powerful enough for something this simple when I play many games at 1080p (I have a GeForce GT 120M btw). Below are my shaders, First Pass #VS #version 330 core uniform mat4 ModelViewMatrix; uniform mat3 NormalMatrix; uniform mat4 MVPMatrix; uniform float scale; layout(location = 0) in vec3 in_Position; layout(location = 1) in vec3 in_Normal; layout(location = 2) in vec2 in_TexCoord; smooth out vec3 pass_Normal; smooth out vec3 pass_Position; smooth out vec2 TexCoord; void main(void){ pass_Position = (ModelViewMatrix * vec4(scale * in_Position, 1.0)).xyz; pass_Normal = NormalMatrix * in_Normal; TexCoord = in_TexCoord; gl_Position = MVPMatrix * vec4(scale * in_Position, 1.0); } #FS #version 330 core uniform sampler2D inSampler; smooth in vec3 pass_Normal; smooth in vec3 pass_Position; smooth in vec2 TexCoord; layout(location = 0) out vec3 outPosition; layout(location = 1) out vec3 outDiffuse; layout(location = 2) out vec3 outNormal; void main(void){ outPosition = pass_Position; outDiffuse = texture(inSampler, TexCoord).xyz; outNormal = pass_Normal; } Second Pass #VS #version 330 core uniform float scale; layout(location = 0) in vec3 in_Position; void main(void){ gl_Position = mat4(1.0) * vec4(scale * in_Position, 1.0); } #FS #version 330 core struct Light{ vec3 direction; }; uniform ivec2 ScreenSize; uniform Light light; uniform sampler2D PositionMap; uniform sampler2D ColorMap; uniform sampler2D NormalMap; out vec4 out_Color; vec2 CalcTexCoord(void){ return gl_FragCoord.xy / ScreenSize; } vec4 CalcLight(vec3 position, vec3 normal){ vec4 DiffuseColor = vec4(0.0); vec4 SpecularColor = vec4(0.0); vec3 light_Direction = -normalize(light.direction); float diffuse = max(0.0, dot(normal, light_Direction)); if(diffuse 0.0){ DiffuseColor = diffuse * vec4(1.0); vec3 camera_Direction = normalize(-position); vec3 half_vector = normalize(camera_Direction + light_Direction); float specular = max(0.0, dot(normal, half_vector)); float fspecular = pow(specular, 128.0); SpecularColor = fspecular * vec4(1.0); } return DiffuseColor + SpecularColor + vec4(0.1); } void main(void){ vec2 TexCoord = CalcTexCoord(); vec3 Position = texture(PositionMap, TexCoord).xyz; vec3 Color = texture(ColorMap, TexCoord).xyz; vec3 Normal = normalize(texture(NormalMap, TexCoord).xyz); out_Color = vec4(Color, 1.0) * CalcLight(Position, Normal); } Is it normal for the GPU to be used that much under the described circumstances? Is it due to poor performance of freeglut? I understand that the problem could be specific to my code, but I can't paste the whole code here, if you need more info, please tell me.

    Read the article

  • zlib memory usage / performance. With 500kb of data.

    - by unixman83
    Is zLib Worth it? Are there other better suited compressors? I am using an embedded system. Frequently, I have only 3MB of RAM or less available to my application. So I am considering using zlib to compress my buffers. I am concerned about overhead however. The buffer's average size will be 30kb. This probably won't get compressed by zlib. Anyone know of a good compressor for extremely limited memory environments? However, I will experience occasional maximum buffer sizes of 700kb, with 500kb much more common. Is zlib worth it in this case? Or is the overhead too much to justify? My sole considerations for compression are RAM overhead of algorithm and performance at least as good as zlib.

    Read the article

  • How and where to store user uploaded files in high traffic web farm scenario website?

    - by Inam Jameel
    i am working on a website which deploy on web farms to serve high traffic. where should i store user uploaded files? is it wise to store uploaded files in the file system of the same website and synchronize these files in all web servers(web farm)? or should i use another server to store all uploaded files in this server to store files in a central location? if separate file server will be a better choice, than how can i pass files from web server to that file server efficiently? or should i upload files directly to that file server?

    Read the article

  • ASP.NET Memory Usage in IIS is FAR greater than in DevEnv. Is this normal?

    - by Tom
    Greetings! I have an ASP.NET app that scrapes data from a handful of external pages, parses the relevant bits and displays them in a table. Total data retrieved is 3-4MB and the resulting page is about 1MB. I am using synchronous WebRequest GetResponse for the retrieval, but the same problem existed using an asynchronous BeginGetResponse/EndGetResponse process. There is no database access, no session storage, no caching, but an in-memory list of about 100 objects (total 1MB of data), plus a good amount of AJAX (AjaxControlToolkit). This issue appears on the very first run of the app, even if I have restarted IIS. The issue: When I run the app on my dev computer, the maximum commit charge is about 1.5GB. The biggest user, measured by Task Manager's VM Size, is WebDev.WebServer.exe (600MB). The app runs perfectly. When I run it on my rent-a-server (IIS 7.5, 1GB RAM), the maximum commit charge is over 3.8GB. The biggest user is w3wp.exe at 2.7GB. IIS grinds to a halt and spits out a timed-out error page. Given my limited server budget and the hope of having multiple simultaneous users, I'm kind of in a panic. Is this normal? If I bump the server RAM up to 4GB, will that be enough? Will multiple users require even more memory? Could the culprit be AJAX or the list of objects? Thanks for any insight you can provide.

    Read the article

  • Understanding the 'High Performance' meaning in Extreme Transaction Processing

    - by kyap
    Despite my previous blogs entries on SOA/BPM and Identity Management, the domain where I'm the most passionated is definitely the Extreme Transaction Processing, commonly called XTP.I came across XTP back to 2007 while I was still FMW Product Manager in EMEA. At that time Oracle acquired a company called Tangosol, which owned an unique product called Coherence that we renamed to Oracle Coherence. Beside this innovative renaming of the product, to be honest, I didn't know much about it, except being a "distributed in-memory cache for Extreme Transaction Processing"... not very helpful still.In general when people doesn't fully understand a technology or a concept, they tend to find some shortcuts, either correct or not, to justify their lack-of understanding... and of course I was part of this category of individuals. And the shortcut was "Oracle Coherence Cache helps to improve Performance". Excellent marketing slogan... but not very meaningful still. By chance I was able to get away quickly from that group in July 2007* at Thames Valley Park (UK), after I attended one of the most interesting workshops, in my 10 years career in Oracle, delivered by Brian Oliver. The biggest mistake I made was to assume that performance improvement with Coherence was related to the response time. Which can be considered as legitimus at that time, because after-all caches help to reduce latency on cached data access, hence reduce the response-time. But like all caches, you need to define caching and expiration policies, thinking about the cache-missed strategy, and most of the time you have to re-write partially your application in order to work with the cache. At a result, the expected benefit vanishes... so, not very useful then?The key mistake I made was my perception or obsession on how performance improvement should be driven, but I strongly believe this is still a common problem to most of the developers. In fact we all know the that the performance of a system is generally presented by the Capacity (or Throughput), with the 2 important dimensions Speed (response-time) and Volume (load) :Capacity (TPS) = Volume (T) / Speed (S)To increase the Capacity, we can either reduce the Speed(in terms of response-time), or to increase the Volume. However we tend to only focus on reducing the Speed dimension, perhaps it is more concrete and tangible to measure, and nicer to present to our management because there's a direct impact onto the end-users experience. On the other hand, we assume the Volume can be addressed by the underlying hardware or software stack, so if we need more capacity (scale out), we just add more hardware or software. Unfortunately, the reality proves that IT is never as ideal as we assume...The challenge with Speed improvement approach is that it is generally difficult and costly to make things already fast... faster. And by adding Coherence will not necessarily help either. Even though we manage to do so, the Capacity can not increase forever because... the Speed can be influenced by the Volume. For all system, we always have a performance illustration as follow: In all traditional system, the increase of Volume (Transaction) will also increase the Speed (Response-Time) as some point. The reason is simple: most of the time the Application logics were not designed to scale. As an example, if you have a while-loop in your application, it is natural to conceive that parsing 200 entries will require double execution-time compared to 100 entries. If you need to "Speed-up" the execution, you can only upgrade your hardware (scale-up) with faster CPU and/or network to reduce network latency. It is technically limited and economically inefficient. And this is exactly where XTP and Coherence kick in. The primary objective of XTP is about designing applications which can scale-out for increasing the Volume, by applying coding techniques to keep the execution-time as constant as possible, independently of the number of runtime data being manipulated. It is actually not just about having an application running as fast as possible, but about having a much more predictable system, with constant response-time and linearly scale, so we can easily increase throughput by adding more hardwares in parallel. It is in general combined with the Low Latency Programming model, where we tried to optimize the network usage as much as possible, either from the programmatic angle (less network-hoops to complete a task), and/or from a hardware angle (faster network equipments). In this picture, Oracle Coherence can be considered as software-level XTP enabler, via the Distributed-Cache because it can guarantee: - Constant Data Objects access time, independently from the number of Objects and the Coherence Cluster size - Data Objects Distribution by Affinity for in-memory data grouping - In-place Data Processing for parallel executionTo summarize, Oracle Coherence is indeed useful to improve your application performance, just not in the way we commonly think. It's not about the Speed itself, but about the overall Capacity with Extreme Load while keeping consistant Speed. In the future I will keep adding new blog entries around this topic, with some sample codes experiences sharing that I capture in the last few years. In the meanwhile if you want to know more how Oracle Coherence, I strongly suggest you to start with checking how our worldwide customers are using Oracle Coherence first, then you can start playing with the product through our tutorial.Have Fun !

    Read the article

  • Win 7 running slowly with low CPU usage and memory

    - by guywhoneedsahand
    I have a relatively new (under 2 yrs old) windows 7 machine. It has 9GB of RAM, and an i7 core CPU (930 @ 2.8GHz w/ 8 CPUs). After about 8 months since a clean install, I noticed my computer was running slowly. I figure it was fragmentation etc, and I did a complete wipe & clean reinstall. However, my problems are somehow persisting. The computer is running painfully slowly (but in leaps and bounds - sometimes it will work fine for 3 hrs, then suddenly freeze up just from clicking the start button). The 'freezes' happen randomly - not during any especially intensive computing. I initially thought something might be eating through my CPU and/or Memory, but Task Manager indicates that neither the CPU or Memory spike. In fact, even during serious lag, CPU usage remains at less than 5% and Memory at ~ 1.5GB. It's beyond me why a fresh install on a powerful machine is performing so poorly... and it certainly is frustrating! What could be causing the poor performance, and what can I do to fix it?

    Read the article

  • High load average, low CPU and IO (Centos 5.7)

    - by Ben
    A Drupal 7 site with CiviCRM, after running smoothly for a year on a 1&1 VPS suddenly became unresponsive. Now pages eventually load, but can take more than a minute. Looking at resource use in Virtuozzo, the load average carries a warning, and has remained above 1. While I understand this isn't particularly high, this is a change from when the site was working. Here is a typical snapshot of top: top - 03:10:32 up 3:21, 1 user, load average: 1.16, 1.22, 1.30 Tasks: 43 total, 1 running, 42 sleeping, 0 stopped, 0 zombie Cpu(s): 0.1%us, 0.1%sy, 0.1%ni, 99.7%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 2097152k total, 1015112k used, 1082040k free, 0k buffers Swap: 0k total, 0k used, 0k free, 0k cached CPU idle level never seems to go much below 70%. wa is virtually always at 0. There appears to be lots of free memory. And here is some vmstat output, again showign wa at 0, plenty of free memory, and an idle CPU: procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------ r b swpd free buff cache si so bi bo in cs us sy id wa st 2 0 0 1100872 0 0 0 0 2783 23672 0 538 1 0 99 0 0 1 0 0 1100872 0 0 0 0 0 16 0 101754 0 0 100 0 0 0 0 0 1100872 0 0 0 0 0 17 0 103133 0 0 100 0 0 0 0 0 1100872 0 0 0 0 0 1 0 102080 0 0 100 0 0 1 0 0 1100872 0 0 0 0 0 6 0 99881 0 0 100 0 0 0 0 0 1100872 0 0 0 0 0 1 0 105187 0 0 100 0 0 I've spoken to 1&1 but they don't have any ideas as to what could be causing the high load average. Instead they suggested an upgrade :) I've looked for processes that might be causing this, examined MySQL showprocesslist, and restarted the container with no result. Does anyone have more troubleshooting suggestions or insights?

    Read the article

  • Virtualbox HTTP load testing, host CPU overload issues

    - by aschuler
    I'm doing HTTP load testing benchmarks (using Apache Benchmark and Siege) on a small Java EE 1.7.0 / Tomcat 7.0.26 application running on a Debian Squeeze 6.0.4 x64 virtualized with Virtualbox 4.1.8. The computer host is Ubuntu 11.10 x64. I've modified those parameters in the Tomcat server.xml : <Connector port="8080" protocol="HTTP/1.1" connectionTimeout="200000" redirectPort="8443" acceptCount="2000" maxThreads="150" minSpareThreads="50" /> The application executed on the server takes around 300ms. This app is running well until a certain amount of concurrent connections like those one : ab -n 500 -c 150 http://xx.xx.xx.xx:8080/myapp/ ab -n 1000 -c 50 http://xx.xx.xx.xx:8080/myapp/ siege -b -c 100 -r 20 http://xx.xx.xx.xx:8080/myapp/ A lot of socket connection timed out happens and this completly overload the host processor (but the CPU load inside the VM is normal). Doing an htop on the host, i can see that the Virtualbox processus is running under 300% CPU and never come down even after the load test is finished. (I've allocated 4 processors to the VM, if I allocate only one processor, CPU load goes under 100%). Restarting Tomcat don't do anything, i'm forced to restart the whole VM. I've tryed to launch those ab/siege commands locally on the VM and everything goes well. I first thought it was related to a linux network limit as explained here: Running some benchmarks using ab, and tomcat starts to really slow down So I've modified those TCP parameters : echo 15 > /proc/sys/net/ipv4/tcp_fin_timeout echo 30 > /proc/sys/net/ipv4/tcp_keepalive_intvl echo 1 > /proc/sys/net/ipv4/tcp_tw_recycle echo 1 > /proc/sys/net/ipv4/tcp_tw_reuse It seems to be better, but it continues to overload the host CPU and output socket connections time out at a certain amount of concurrent connections. I'm wondering if this is not related to how Virtualbox handles external concurrent connections.

    Read the article

  • mysqld causes high CPU load

    - by Radu
    My mysqld goes to use 99.9% of CPU for variable time (between 2 - 20 minutes), and then goes back to normal 0.1% - 5%. Checked processlist: all is normal, 1 to 20 inserts or updates that last 2 to 5 sec, and about 20 process that are in Sleep Mode (maybe because the scripts don't close the mysql connection, but are they are closed in about 5 - 10 secs, I didn't make the scripts :P but the server was running fine the last 2 years, since is was made): | 15375 | root | localhost | stoc | Query | 0 | NULL | show processlist | | 79480 | pppoe | localhost | pppoe | Sleep | 4 | NULL | NULL | | 79481 | pppoe | localhost | pppoe | Sleep | 4 | NULL | NULL | | 79482 | pppoe | localhost | pppoe | Sleep | 4 | NULL | NULL | | 79483 | pppoe | localhost | pppoe | Query | 0 | init | UPDATE acc SET InputOctets="0", OutputOctets="0", InputPackets="unknown", OutputPackets="User | | 79484 | pppoe | localhost | pppoe | Sleep | 5 | NULL | NULL | | 79485 | pppoe | localhost | pppoe | Sleep | 5 | NULL | NULL | | 79486 | pppoe | localhost | pppoe | Sleep | 5 | NULL | NULL Checked raid, seemns OK: [root@db2]# cat /proc/mdstat Personalities : [raid5] [raid4] [raid1] md0 : active raid1 sdd1[3] sdc1[2] sdb1[0] sda1[1] 136448 blocks [4/4] [UUUU] md1 : active raid5 sdd2[3] sdc2[2] sdb2[0] sda2[1] 12023808 blocks level 5, 256k chunk, algorithm 2 [4/4] [UUUU] md3 : active raid5 sda4[1] sdd4[3] sdc4[2] sdb4[0] 203647488 blocks level 5, 256k chunk, algorithm 2 [4/4] [UUUU] md2 : active raid5 sda3[1] sdd3[3] sdc3[2] sdb3[0] 24024576 blocks level 5, 256k chunk, algorithm 2 [4/4] [UUUU] unused devices: <none> [root@db2]# top sees my mysqld cpu load, but nothing else seems to be wrong: [root@db2]# top top - 17:56:05 up 7 days, 3:55, 3 users, load average: 32.93, 24.72, 22.70 Tasks: 75 total, 4 running, 71 sleeping, 0 stopped, 0 zombie Cpu(s): 63.4% us, 36.6% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.0% hi, 0.0% si, 0.0% st Mem: 1988824k total, 1304776k used, 684048k free, 99588k buffers Swap: 12023800k total, 0k used, 12023800k free, 951028k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 5754 mysql 19 0 236m 57m 5108 R 99.9 2.9 21:58.76 mysqld 1 root 16 0 7216 700 580 S 0.0 0.0 0:00.39 init 2 root RT 0 0 0 0 S 0.0 0.0 0:00.00 migration/0 Repaired all mysql databases, reindexed raid ... I'm running out of ideeas ... Anyone has an ideea what can go wrong with this server ? Thank you

    Read the article

  • Apache/Passenger and cpulimit

    - by Dave Smylie
    I run a ruby on rails site that processes email - the email is dumped directly into the web app via a POST from postfix. At times I can get a burst of email coming in causing a prolonged surge in CPU usage making my VPS provider understandable unhappy with me. These emails don't need to be processed in a timely manner - they just need to be (eventually) processed. Obviously I can't just nice the process as that only looks at the cpu usage on my VPS and can't take into account the cpu usage on the other VPS's. I have found a utility called cpulimit that will you put hard limits on cpu usage for a particular process. (eg 20%). This seems ideal for this purpose, but I can't work out to integrate with apache/passenger. Passenger starts up a ruby process for each server and restarts them periodically. Each time the pid will change. Cpulimit needs to be given a pid number for it to act on. Anyone got any ideas how I could get passenger to fire off a call this command when it's starting up this particular virtual host?

    Read the article

  • OpenVPN Server - CPU is pegged out

    - by ericl42
    Hello, I am configuring OpenVPN to act as a SSL tunnel for a remote location. I have OpenVPN1 at our current location acting as a server then OpenVPN2 at the other location that is acting as a client but is also acting as a DHCP server to machines behind it so they are basically connected to the local LAN. Everything is set up fine and I can talk from location A to location B with no problems like everyone is local. I am however having some performance issues. OpenVPN1 CPU is pegged to 100% the entire time I am copying or doing any type of activity through the tunnel. I expect some CPU usage going up but nothing like this. It's really killing my performance. OpenVPN1 is running in ESX right now with 2 gig RAM and 4 procs with unlimited bursting capacity. I am using AES-192 encryption with a 1024 key. Any idea how I can get my CPU down on OpenVPN1 and my download/upload speeds higher between the tunnel? Thanks. edit: Turning down the logging helped boost the throughput a little bit, but I am still fairly shy of where I believe I should be. Also I am still maxed out on the CPU. Does anyone have any ideas? I am really stuck on this. Thanks.

    Read the article

  • Extremely high mysqld CPU usage with no active queries

    - by RadarNyan
    I have a VPS running Ubuntu 12.04 LTS with LEMP stack, followed the guide from Linode Library (since I'm using a Linode) to setup, and everything worked fine until now. I don't know what's wrong, but my CPU usage just goes up since a week ago. Today things getting really bad - I got 74% CPU usage so I went check and found that mysqld taking too much CPU usage (somewhere around 30% ~ 80%) So I did some Google Search, tried disable InnoDB, restart mysql, reset ntp / system clock (Isn't this bug supposed to happen more than a year ago?!) and reboot my VPS, nothing helped. Even with mysql processlist empty, I still get mysqld CPU usage very high. I don't know what I missed and have totally no idea, any advice would be appreciated. Thanks in advance. Update: I got these from running "strace mysqld" write(2, "InnoDB: Unable to lock ./ibdata1"..., 44) = 44 write(2, "InnoDB: Check that you do not al"..., 115) = 115 select(0, NULL, NULL, NULL, {1, 0}^[[A^[[A) = 0 (Timeout) fcntl64(3, F_SETLK64, {type=F_WRLCK, whence=SEEK_SET, start=0, len=0}, 0xbfa496f8) = -1 EAGAIN (Resource temporarily unavailable) hum... I did tried to disable InnoDB and it didn't fix this problem. Any idea? Update2: # ps -e | grep mysqld 13099 ? 00:00:20 mysqld then use "strace -p 13099", the following lines appears repeatedly: fcntl64(12, F_GETFL) = 0x2 (flags O_RDWR) fcntl64(12, F_SETFL, O_RDWR|O_NONBLOCK) = 0 accept(12, {sa_family=AF_FILE, NULL}, [2]) = 14 fcntl64(12, F_SETFL, O_RDWR) = 0 getsockname(14, {sa_family=AF_FILE, path="/var/run/mysqld/mysqld.sock"}, [30]) = 0 fcntl64(14, F_SETFL, O_RDONLY) = 0 fcntl64(14, F_GETFL) = 0x2 (flags O_RDWR) setsockopt(14, SOL_SOCKET, SO_RCVTIMEO, "\36\0\0\0\0\0\0\0", 8) = 0 setsockopt(14, SOL_SOCKET, SO_SNDTIMEO, "<\0\0\0\0\0\0\0", 8) = 0 fcntl64(14, F_SETFL, O_RDWR|O_NONBLOCK) = 0 setsockopt(14, SOL_IP, IP_TOS, [8], 4) = -1 EOPNOTSUPP (Operation not supported) futex(0xb786a584, FUTEX_WAKE_OP_PRIVATE, 1, 1, 0xb786a580, {FUTEX_OP_SET, 0, FUTEX_OP_CMP_GT, 1}) = 1 futex(0xb7869998, FUTEX_WAKE_PRIVATE, 1) = 1 poll([{fd=10, events=POLLIN}, {fd=12, events=POLLIN}], 2, -1) = 1 ([{fd=12, revents=POLLIN}]) er... now I totally don't get it x_x help

    Read the article

  • windows 95 virtualization and CPU idle

    - by brtlvrs
    We are in the situation that we need to migrate our windows 95 systems (i know, that is so last century ). There is no hardware for a reasonable price available to run our windows 95 systems. So we are extending the overdue lifetime of it, by making them virtual. Now we see that VMware reports 100% CPU utilization of a windows 95 VM. This is because windows 95 doesn't know how to manage a CPU. For this reason software like rain or Waterfall, or CPUCool are introduced. The they are sending a HLT instruction to the CPU. This causes the CPU to halt and wait for new triggers to work. I've tested the mentioned programs in the VM, but they don't work, but generate errors. anyone has a valid workaround, solution ?? btw. I know that the best solutiuon is to replace the windows95 with windows XP. But in our situation that will take at least 5 years. Our windows 95 systems are running factory process control software.....

    Read the article

  • UW-IMAP server, high load for one user

    - by Bruce Garlock
    We have been experiencing a very strange anomaly, with one specific user with our UW-IMAP server. We have about 75 users using the server, and one particular user, who is in about the middle as far as used storage keeps having issues with slow speed. Most of our users all use Thunderbird 2, or Thunderbird 3. Mostly 2, because of the performance issues we have had with 3. This user was on 3, and I downgraded him to 2. The performance has gotten better, but according to the imapd processes on the server, his username is using the most CPU % and CPU time. I've already done all the usual T/S'ing: Started profile from scratch, compacted folders, re-indexed, newer faster computer, etc.. Still, this users' imapd process is always using the most CPU on the server. For troubleshooting, we setup another user which has more usage, folders, etc.. than he does, but we don't see the users process taking up most of the CPU with the imapd process. So, it almost sounds like a particular email may be the culprit, but how can we find it, if thats the problem? This has been going on for a while, and he is a management person, so his patience is about to end. Does anyone have any ideas?

    Read the article

  • Random server lag, no CPU/mem/pagefile usage

    - by Kev
    We have a fairly new server running Windows 2003 SP2, and the past few days we've noticed random slowdowns. When I'm logged into the server over remote desktop while this is happening, or if I'm physically sitting at the server logged in, suddenly everything becomes extremely laggy. Any UI element I try to interact with takes upwards of ten seconds to react, and then responds very slowly. Then a minute later everything is quite snappy again. During this, I have Task Manager minimized to the tray, and there's no CPU usage. I open it up right after this happens, and there's very little CPU usage on the graph, and no memory or pagefile usage above normal. (Normal being 1.5 GB free in the case of memory.) This is what I see logged into the server, and then users start calling saying things are slow, timing out, and failing--anything to do with our server. No events in the Event Viewer around the times this happens. The context I'm working in (last thing I clicked, etc.) seems different every time--different programs active, different combinations of programs open. Never anything particularly stressful (like adding an event entry to a Cobian Backup configuration, or editing text in TextPad, which has been exceptionally stable in my extensive usage of it.) I would've thought it was just the server, but a family member's home PC (entirely separate) running WinXPSP3 had the same thing happen to it last night a few times. Is this some new behaviour introduced by the latest Windows Updates? Either way, where do I even start to look when nothing seems to be chewing up resources?

    Read the article

  • Getting a "CPU over temperature error", but temperatures seem to be normal

    - by Luis Parker
    I built a PC with the following parts: CPU: i5-2500 Motherboard: Asus P8H67-M EVO rev 3.0 RAM: G-Skill Ripjaws Series 8GB (2 x 4GB) DDR3 1333 Video: GTX 560-Ti 1GB I used a crappy (but functioning) old case and a 500W Powercooler supply. The rig also includes 3 HDs and a DVD-RW. Whenever I push the system mildly it resets right away (looks more like power off/power on) and gives me a "CPU over temperature error". However, BIOS always reports <65ºC for the CPU at the moment of the reset, same as Core Temp and RealTemp. This is RealTemp's log since launching BF3 until the reset (just over 1 minute, as you can see) Just to be sure I've checked the CPU cooler and re-applied the thermal paste twice, but nothing changed. I'm not overclocking at all. What am I missing here? Could it be that the old power supply is generating this error? Maybe the mobo isn't reporting temperatures correctly? I don't have a clue on how to troubleshoot this, help Thanks in advance!

    Read the article

  • Is it possible that a single-threaded program is executed simultaneously on more than one CPU core?

    - by Wolfgang Plaschg
    When I run a single-threaded program that i have written on my quad core Intel i can see in the Windows Task Manager that actually all four cores of my CPU are more or less active. One core is more active than the other three, but there is also activity on those. There's no other program (besided the OS kernel of course) running that would be plausible for that activitiy. And when I close my program all activity an all cores drops down to nearly zero. All is left is a little "noise" on the cores, so I'm pretty sure all the visible activity comes directly or indirectly (like invoking system routines) from my program. Is it possible that the OS or the cores themselves try to balance some code or execution on all four cores, even it's not a multithreaded program? Do you have any links that documents this technique? Some infos to the program: It's a console app written in Qt, the Task Manager states that only one thread is running. Maybe Qt uses threads, but I don't use signals or slots, nor any GUI. Link to Task Manager screenshot: http://img97.imageshack.us/img97/6403/taskmanager.png This question is language agnostic and not tied to Qt/C++, i just want to know if Windows or Intel do to balance also single-threaded code on all cores. If they do, how does this technique work? All I can think of is, that kernel routines like reading from disk etc. is scheduled on all cores, but this won't improve performance significantly since the code still has to run synchronous to the kernel api calls. EDIT Do you know any tools to do a better analysis of single and/or multi-threaded programs than the poor Windows Task Manager?

    Read the article

  • Current trends in Random Access Memory

    - by Nutel
    As I know for now because of laws of Physics there will be not any tangible improvements in CPU cycles per second for the nearest future. However because of Von Neumann bottleneck it seems to not be an issue for non-server applications. So what about RAM, is there any upcoming technologies that promise to improve memory speed or we are stack with the current situation till quantum computers will come out from labs?

    Read the article

  • When will we get Sandy Bridge support?

    - by Gu3miles
    I'd like to build a machine running Ubuntu and using the new Sandy bridge (i5 2500 specifically) intel cpu's. I heard that 10.10 doesn't have the best support (or simply won't work) with the new cpus and 1155 mobos. I'm assuming that 11.04 will have support but do the alpha's have it already? Or will I have to wait? Also, I plan to use the onboard Intel graphics (H67 mobo), will there be support for this or will I still need to use a graphic's card?

    Read the article

  • How to split up levels? (cocos2d,box2d,iphone) to save CPU and memory?

    - by cocos2dbeginner
    Hi, so I'm going to create large levels. But there's a problem: There's much unseen space (it's a jump'n run like mario bros.) and this will use memory + cpu. so how could I split up my levels? I'm using Box2D+ cocos2d for iphone. Any ideas? Mayby just set the visible property to NO? But it would be still in the memory :(. But what with the box2d bodies? Destroy and recreate them would be to heavy for the FPS, because I have physics built in which should not be recreated. Should I make fix points where i want to split the level up, than if the player is 200 px away it should preload it. and if the player is 200 px away from the last part of the level I unload it. But there would be the problem with the physics, because on the start of the level it has a unique movement and later if i destroy and recreate it it would do the same. but i don't want that. other ideas?

    Read the article

  • ACPI 30k+ interrupts per second

    - by ultimatebuster
    My computer is currently doing something like 30k+ interrupts per second. This causes battery issues as well as the CPU being always stuck at the highest speed, making it very hot (overheating.. 75C at IDLE) Powertop info: 99.4% (32052.4) [acpi] <interrupt> 0.3% (115.9) [iwlagn] <interrupt> Just to clarify this started suddenly a couple of days ago. It didn't have this issue before (though I had a tonne of headaches with Linux on my computer in general..)

    Read the article

< Previous Page | 35 36 37 38 39 40 41 42 43 44 45 46  | Next Page >