Search Results

Search found 3140 results on 126 pages for 'stencil buffer'.

Page 10/126 | < Previous Page | 6 7 8 9 10 11 12 13 14 15 16 17  | Next Page >

• linux kernel buffer memory is zero

  - by user64772

  Hi all. There are one qestion that i can`t find in google. I have many linux boxes mostly with SLES or openSUSE, diffrent versions and kernels. On some of them i faced with slow oracle transactions problem. It time to time problem and when i log in the box on that time i see that oracle blocked in kernel function sync_page # while :; do ps axo stat,pid,cmd,wchan | egrep '^D|^R'; echo --; sleep 5; done D 3483 hald-addon-storage: polling ide_do_drive_cmd Ds 4635 ora_dbw0_orcl sync_page Ds 4637 ora_lgwr_orcl sync_page Ds 4639 ora_ckpt_orcl sync_page D 11210 oracleorcl (LOCAL=NO) sync_page D 12457 [smtpd] sync_page R+ 12458 ps axo stat,pid,cmd,wchan - -- Ds 4635 ora_dbw0_orcl sync_page Ds 4637 ora_lgwr_orcl sync_page Ds 4639 ora_ckpt_orcl sync_page D 11210 oracleorcl (LOCAL=NO) sync_page R+ 12501 ps axo stat,pid,cmd,wchan - -- Ds 4635 ora_dbw0_orcl sync_page Ds 4637 ora_lgwr_orcl sync_page Ds 4639 ora_ckpt_orcl sync_page D 11210 oracleorcl (LOCAL=NO) sync_page R+ 12535 ps axo stat,pid,cmd,wchan - -- Ds 4635 ora_dbw0_orcl sync_page Ds 4637 ora_lgwr_orcl sync_page Ds 4639 ora_ckpt_orcl sync_page D 11210 oracleorcl (LOCAL=NO) sync_page R+ 12570 ps axo stat,pid,cmd,wchan - -- so i think that box is run out of memory for disk buffers but memry is fine total used free shared buffers cached Mem: 4149084 3994552 154532 0 0 2424328 -/+ buffers/cache: 1570224 2578860 Swap: 3148700 750696 2398004 i think that this is the problem, buffer is zero and we must write directly to disk, but why buffer is zero ? - i try to google it and find nothing - is anyone can help ?

  Read the article

• How to detect the character encoding of a text file?

  - by Cédric Boivin

  I try to detect which character encoding is used in my file. I try with this code to get the standard encoding public static Encoding GetFileEncoding(string srcFile) { // *** Use Default of Encoding.Default (Ansi CodePage) Encoding enc = Encoding.Default; // *** Detect byte order mark if any - otherwise assume default byte[] buffer = new byte[5]; FileStream file = new FileStream(srcFile, FileMode.Open); file.Read(buffer, 0, 5); file.Close(); if (buffer[0] == 0xef && buffer[1] == 0xbb && buffer[2] == 0xbf) enc = Encoding.UTF8; else if (buffer[0] == 0xfe && buffer[1] == 0xff) enc = Encoding.Unicode; else if (buffer[0] == 0 && buffer[1] == 0 && buffer[2] == 0xfe && buffer[3] == 0xff) enc = Encoding.UTF32; else if (buffer[0] == 0x2b && buffer[1] == 0x2f && buffer[2] == 0x76) enc = Encoding.UTF7; else if (buffer[0] == 0xFE && buffer[1] == 0xFF) // 1201 unicodeFFFE Unicode (Big-Endian) enc = Encoding.GetEncoding(1201); else if (buffer[0] == 0xFF && buffer[1] == 0xFE) // 1200 utf-16 Unicode enc = Encoding.GetEncoding(1200); return enc; } My five first byte are 60, 118, 56, 46 and 49. Is there a chart that shows which encoding matches those five first bytes?

  Read the article

• Actionscript: NetStream stutters after buffering.

  - by meandmycode

  Using NetStream to stream content from http, I've noticed that esp with certain exported h264's, if the player encounters an empty buffer, it will stop and buffer to the requested length (as expected). However once the buffer is full, the playback doesn't resume, but instead jumps ahead, as such- instantly playing the buffered duration in a brief moment, and thusly triggering an empty buffer again.. this will then continue over and over. Presumably when the netstream pauses to buffer, the playhead position continues, and the player is attempting to snap to that position on resume- however given it could take 5 seconds to build a 2 second buffer- it ends up with a useless buffer again.. (this is an assumption) I've attempted to work around this by listening for an empty buffer netstatus event, pausing the stream, and at the same time setting up a loop to check the current buffer length vs the requested buffer length.. and resuming once the buffer length is greater than or equal to the requested buffer.. however this causes problems when there isn't enough of the video remaining.. for example, a 10 second buffer with only 5 seconds remaining, the loop just sits there waiting for a buffer length of 10 seconds when theres only 5 left... You would think that you could simply check which was smaller, the time left or the requested buffer length.. however the times flash gives are not accurate.. If you add the net streams current time index, plus the buffered time, the total is not the entire duration of the movie (when at the end).. it is close but not the same. This brings me back to the original problem, and if there is another way to fix this, clearly flash knows when the buffer is ready, so how can i get flash pause when it buffers, and resume once the buffer is ready? currently it doesn't.. it pauses and then once the buffer is full- it plays the entire buffered content in about .1 of a second. Thanks in advance, Stephen.

  Read the article

• Asynchronous readback from opengl front buffer using multiple PBO's

  - by KillianDS

  I am developing an application that needs to read back the whole frame from the front buffer of an openGL application. I can hijack the application's opengl library and insert my code on swapbuffers. At the moment I am successfully using a simple but excruciating slow glReadPixels command without PBO's. Now I read about using multiple PBO's to speed things up. While I think I've found enough resources to actually program that (isn't that hard), I have some operational questions left. I would do something like this: create a series (e.g. 3) of PBO's use glReadPixels in my swapBuffers override to read data from front buffer to a PBO (should be fast and non-blocking, right?) Create a seperate thread to call glMapBufferARB, once per PBO after a glReadPixels, because this will block until the pixels are in client memory. Process the data from step 3. Now my main concern is of course in steps 2 and 3. I read about glReadPixels used on PBO's being non-blocking, will this be an issue if I issue new opengl commands after that very fast? Will those opengl commands block? Or will they continue (my guess), and if so, I guess only swapbuffers can be a problem, will this one stall or will glReadPixels from front buffer be many times faster than swapping (about each 15-30ms) or, worst case scenario, will swapbuffers be executed while glReadPixels is still reading data to the PBO? My current guess is this logic will do something like this: copy FRONT_BUFFER - generic place in VRAM, copy VRAM-RAM. But I have no idea which of those 2 is the real bottleneck and more, what the influence on the normal opengl command stream is. Then in step 3. Is it wise to do this asynchronously in a thread separated from normal opengl logic? At the moment I think not, It seems you have to restore buffer operations to normal after doing this and I can't install synchronization objects in the original code to temporarily block those. So I think my best option is to define a certain swapbuffer delay before reading them out, so e.g. calling glReadPixels on PBO i%3 and glMapBufferARB on PBO (i+2)%3 in the same thread, resulting in a delay of 2 frames. Also, when I call glMapBufferARB to use data in client memory, will this be the bottleneck or will glReadPixels (asynchronously) be the bottleneck? And finally, if you have some better ideas to speed up frame readback from GPU in opengl, please tell me, because this is a painful bottleneck in my current system. I hope my question is clear enough, I know the answer will probably also be somewhere on the internet but I mostly came up with results that used PBO's to keep buffers in video memory and do processing there. I really need to read back the front buffer to RAM and I do not find any clear explanations about performance in that case (which I need, I cannot rely on "it's faster", I need to explain why it's faster). Thank you

  Read the article

• Keyboard "type ahead" in CRUD web apps?

  - by user61852

  In some data entry contexts, I've seen data typists, type really fast and know so well the app they use, and have a mechanic quality in their work so that they can "type ahead", ie continue typing and "tab-bing" and "enter-ing" faster than the display updates, so that in many occasions they are typing in the data for the next form before it draws itself. Then when this next entry form appears, their keystrokes fill the text boxes and they continue typing, selecting etc. In contexts like this, this speed is desirable, since this persons are really productive. I think this "type ahead of time" is only possible in desktop apps, but I may be wrong. My question is whether this way of handling the keyboard buffer (which in desktop apps require no extra programming) is achievable in web apps, or is this impossible because of the way web apps work, handle sessions, etc (network latency and the overhead of generating new web pages ) ?

  Read the article

• OpenGL problem with FBO integer texture and color attachment

  - by Grieverheart

  In my simple renderer, I have 2 FBOs one that contains diffuse, normals, instance ID and depth in that order and one that I use store the ssao result. The textures I use for the first FBO are RGB8, RGBA16F, R32I and GL_DEPTH_COMPONENT32F for the depth. For the second FBO I use an R16F texture. My rendering process is to first render to everything I mentioned in the first FBO, then bind depth and normals textures for reading for the ssao pass and write to the second FBO. After that I bind the second FBO's texture for reading in my blur shader and bind the first FBO for writing. What I intend to do is to write the blurred ssao value to the alpha component of the Normals texture. Here are where the problems start. First of all, I use shading language 3.3, which my graphics card does support. I manage ouputs in my shaders using layout(location = #). Now, the normals texture should be bound to color attachment 1, but when I use 1, it seems to write to my diffuse texture which should be in color attachment 0. When I instead use layout(location = 0), it gets correctly written to my normals texture. Besides this, my instance ID texture also gets resets after running the blur shader which is weird because if I use a float texture and write to it instanceID / nInstances, the texture doesn't get reset after the blur shader has ran. Here is how I prepare my first FBO: bool CGBuffer::Init(unsigned int WindowWidth, unsigned int WindowHeight){ //Create FBO glGenFramebuffers(1, &m_fbo); glBindFramebuffer(GL_DRAW_FRAMEBUFFER, m_fbo); //Create gbuffer and Depth Buffer Textures glGenTextures(GBUFF_NUM_TEXTURES, &m_textures[0]); glGenTextures(1, &m_depthTexture); //prepare gbuffer for(unsigned int i = 0; i < GBUFF_NUM_TEXTURES; i++){ glBindTexture(GL_TEXTURE_2D, m_textures[i]); if(i == GBUFF_TEXTURE_TYPE_NORMAL) glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA16F, WindowWidth, WindowHeight, 0, GL_RGBA, GL_FLOAT, NULL); else if(i == GBUFF_TEXTURE_TYPE_DIFFUSE) glTexImage2D(GL_TEXTURE_2D, 0, GL_RGB8, WindowWidth, WindowHeight, 0, GL_RGB, GL_FLOAT, NULL); else if(i == GBUFF_TEXTURE_TYPE_ID) glTexImage2D(GL_TEXTURE_2D, 0, GL_R32I, WindowWidth, WindowHeight, 0, GL_RED_INTEGER, GL_INT, NULL); else{ std::cout << "Error in FBO initialization" << std::endl; return false; } glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_COLOR_ATTACHMENT0 + i, GL_TEXTURE_2D, m_textures[i], 0); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP); } //prepare depth buffer glBindTexture(GL_TEXTURE_2D, m_depthTexture); glTexImage2D(GL_TEXTURE_2D, 0, GL_DEPTH_COMPONENT32F, WindowWidth, WindowHeight, 0, GL_DEPTH_COMPONENT, GL_FLOAT, NULL); glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_TEXTURE_2D, m_depthTexture, 0); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST); glTexParameteri (GL_TEXTURE_2D, GL_TEXTURE_COMPARE_MODE, GL_NONE); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP); glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP); GLenum DrawBuffers[] = {GL_COLOR_ATTACHMENT0, GL_COLOR_ATTACHMENT1, GL_COLOR_ATTACHMENT2}; glDrawBuffers(GBUFF_NUM_TEXTURES, DrawBuffers); GLenum Status = glCheckFramebufferStatus(GL_FRAMEBUFFER); if(Status != GL_FRAMEBUFFER_COMPLETE){ std::cout << "FB error, status 0x" << std::hex << Status << std::endl; return false; } //Restore default framebuffer glBindFramebuffer(GL_FRAMEBUFFER, 0); return true; } where I use an enum defined as, enum GBUFF_TEXTURE_TYPE{ GBUFF_TEXTURE_TYPE_DIFFUSE, GBUFF_TEXTURE_TYPE_NORMAL, GBUFF_TEXTURE_TYPE_ID, GBUFF_NUM_TEXTURES }; Am I missing some kind of restriction? Does the color attachment of the FBO's textures somehow gets reset i.e. I'm using a re-size function which re-sizes the textures of the FBO but should I perhaps call glFramebufferTexture2D again too? EDIT: Here is the shader in question: #version 330 core uniform sampler2D aoSampler; uniform vec2 TEXEL_SIZE; // x = 1/res x, y = 1/res y uniform bool use_blur; noperspective in vec2 TexCoord; layout(location = 0) out vec4 out_AO; void main(void){ if(use_blur){ float result = 0.0; for(int i = -1; i < 2; i++){ for(int j = -1; j < 2; j++){ vec2 offset = vec2(TEXEL_SIZE.x * i, TEXEL_SIZE.y * j); result += texture(aoSampler, TexCoord + offset).r; // -0.004 because the texture seems to be a bit displaced } } out_AO = vec4(vec3(0.0), result / 9); } else out_AO = vec4(vec3(0.0), texture(aoSampler, TexCoord).r); }

  Read the article

• How do I use unpackHalf2x16?

  - by user1032861

  I'm trying to use (un)packHalf2x16, without success so far. I'm drawing with: glVertexAttribIPointer(0, 2, GL_UNSIGNED_INT, 0, 0); glEnableVertexAttribArray(0); glBindBuffer(GL_ARRAY_BUFFER, vbo); glDrawArrays(GL_POINTS, 0, n_points); glDisableVertexAttribArray(0); and on the shader #version 330 core #extension GL_ARB_shading_language_packing : require in uvec2 A0; // (...) vec4 t = vec4(unpackHalf2x16(A0.x), unpackHalf2x16(A0.y)); But nothing gets drawn. I'm pretty sure buffer's content is right, and if I use vec4 t = vec4(0); I can see it's working properly. How is this packing / unpacking thing supposed to work? I can't find any example.

  Read the article

• Multiple ( V- / I- ) Buffers, is it sane?

  - by Techie

  Currently I am developing an RTS game using XNA ( / ANX.Framework ). There is one thing bothers me. I am not sure in what way or how to organise Buffers. Should I use a new Vertexbuffer for any object ( e.g. a Char, a Table, an model ) or is it better to use ONE HUGE/ BIG Buffer to store any geometry in? I am still new to 3D Programming though I finished yet couple games using DirectX 9. Well, I hope this question qin't a duplicate and I appreciate any answer leading me into the right direction.

  Read the article

• AudioQueue ate my buffer (first 15 milliseconds of it)

  - by iter

  I am generating audio programmatically. I hear gaps of silence between my buffers. When I hook my phone to a scope, I see that the first few samples of each buffer are missing, and in their place is silence. The length of this silence varies from almost nothing to as much as 20 ms. My first thought is that my original callback function takes too much time. I replace it with the shortest one possible--it re-renqueues the same buffer over and over. I observe the same behavior. AudioQueueRef aq; AudioQueueBufferRef aq_buffer; AudioStreamBasicDescription asbd; void aq_callback (void *aqData, AudioQueueRef inAQ, AudioQueueBufferRef inBuffer) { OSStatus s = AudioQueueEnqueueBuffer(aq, aq_buffer, 0, NULL); } void aq_init(void) { OSStatus s; asbd.mSampleRate = AUDIO_SAMPLES_PER_S; asbd.mFormatID = kAudioFormatLinearPCM; asbd.mFormatFlags = kLinearPCMFormatFlagIsSignedInteger | kAudioFormatFlagIsPacked; asbd.mBytesPerPacket = 1; asbd.mFramesPerPacket = 1; asbd.mBytesPerFrame = 1; asbd.mChannelsPerFrame = 1; asbd.mBitsPerChannel = 8; asbd.mReserved = 0; int PPM_PACKETS_PER_SECOND = 50; // one buffer is as long as one PPM frame int BUFFER_SIZE_BYTES = asbd.mSampleRate/PPM_PACKETS_PER_SECOND*asbd.mBytesPerFrame; s = AudioQueueNewOutput(&asbd, aq_callback, NULL, CFRunLoopGetCurrent(), kCFRunLoopCommonModes, 0, &aq); s = AudioQueueAllocateBuffer(aq, BUFFER_SIZE_BYTES, &aq_buffer); // put samples in the buffer buffer_data(my_data, aq_buffer); s = AudioQueueStart(aq, NULL); s = AudioQueueEnqueueBuffer(aq, aq_buffer, 0, NULL); }

  Read the article

• How can I pare down Vim's buffer list to only include active buffers

  - by nelstrom

  How can I pare down my buffer list to only include buffers that are currently open in a window/tab? When I've been running Vim for a long time, the list of buffers revealed by the :ls command is too large to work with. Ideally, I would like to delete all of the buffers which are not currently visible in a tab or window by running a custom command such as :Only. Can anybody suggest how to achieve this? It looks like the :bdelete command can accept a list of buffer numbers, but I'm not sure how to translate the output from :ls to a format that can be consumed by the :bdelete command. Any help would be appreciated. Clarification Lets say that in my Vim session I have opened 4 files. The :ls command outputs: :ls 1 a "abc.c" 2 h "123.c" 3 h "xyz.c" 4 a "abc.h" Buffer 1 is in the current tab, and and buffer 4 is in a separate tab, but buffers 2 and 3 are both hidden. I would like to run the command :Only, and it would wipe buffers 2 and 3, so the :ls command would output: :ls 1 a "abc.c" 4 a "abc.h" This example doesn't make the proposed :Only command look very useful, but if you have a list of 40 buffers it would be very welcome.

  Read the article

• mmap only needed pages of kernel buffer to user space

  - by axeoth

  See also this answer: http://stackoverflow.com/a/10770582/1284631 I need something similar, but without having to allocate a buffer: the buffer is large, in theory, but the user space program only needs to access some parts of it (it mocks some registers of a hardware). As I cannot allocate with vmalloc_user() such a large buffer (kernel 32 bit, in embedded environment, no swap...), I followed the same approach as in the quoted answer, trying to allocate only those pages that are really requested by the user space. So: I use a my_mmap() function for the device file (actually, is the .mmap field of a struct uio_info) to set up the fields of the vma, then, in the vm_area_struct's .fault field (also named my_fault()), I should return a page. except that: In the my_fault() method of vm_area_struct, I cannot obtain a page through: vmf->page=vmalloc_to_page(my_buf + (vmf->pgoff << PAGE_SHIFT)); since there is no allocated buffer: my_buf = vmalloc_user(MY_BUF_SIZE); fails with "allocation failed: out of vmalloc space - use vmalloc= to increase size." (and there is no room or swap to increase that vmalloc= parameter). So, I would need to get a page from the kernel and fill the vmf->page field. How to allocate a page (I assume that the offset of the page is known, as it is vm->pgoff). What base memory should I use instead of my_buf? PS: I also did set up the vma->flags |= VM_NORESERVE; (in the my_mmap()), but not sure if it helps. Is there any vmalloc_user_unreserved()-like function? (let's say, lazy allocation) Also, writing 1 to /proc/sys/vm/overcommit_memory and large values (eg 500) to /proc/sys/vm/overcommit_ratio before trying to my_buf=vmalloc_user(<large_size>) didn't work.

  Read the article

• std::ifstream buffer caching

  - by ledokol

  Hello everybody, In my application I'm trying to merge sorted files (keeping them sorted of course), so I have to iterate through each element in both files to write the minimal to the third one. This works pretty much slow on big files, as far as I don't see any other choice (the iteration has to be done) I'm trying to optimize file loading. I can use some amount of RAM, which I can use for buffering. I mean instead of reading 4 bytes from both files every time I can read once something like 100Mb and work with that buffer after that, until there will be no element in buffer, then I'll refill the buffer again. But I guess ifstream is already doing that, will it give me more performance and is there any reason? If fstream does, maybe I can change size of that buffer? added My current code looks like that (pseudocode) // this is done in loop int i1 = input1.read_integer(); int i2 = input2.read_integer(); if (!input1.eof() && !input2.eof()) { if (i1 < i2) { output.write(i1); input2.seek_back(sizeof(int)); } else input1.seek_back(sizeof(int)); output.write(i2); } } else { if (input1.eof()) output.write(i2); else if (input2.eof()) output.write(i1); } What I don't like here is seek_back - I have to seek back to previous position as there is no way to peek 4 bytes too much reading from file if one of the streams is in EOF it still continues to check that stream instead of putting contents of another stream directly to output, but this is not a big issue, because chunk sizes are almost always equal. Can you suggest improvement for that? Thanks.

  Read the article

• Network Error: no buffer space available

  - by braindump

  After some time of running fine, one of our Windows XP SP3 machines does not open some(!) new TCP/IP connections anymore. Putty says "Network Error: no buffer space available", IE won't open any new connections but e.g. network drive mappings still work, even new ones can be established. netstat does not show more open connections that usual, ping and dns lookups work fine. Any hints?

  Read the article

• ActiveSync / Exchange 2007 password expiration buffer on device

  - by Matt Hamende

  I'm trying to determine if there is any buffer of time from the time a password expires in AD to the time that users would stop receiving email on their mobile devices our setup is Exchange 2007 ActiveSync DC's are Server 2008 R2 primarily Android shop, with maybe a few iOS devices I've heard some rumors of people still receiving email after their password expired / changed on the domain, just want to see if anyone else has ever heard of this. Did a bit more reading, read about Token Cache in IIS 7.0 and 15min lagtime, still would like to hear any thoughts about this.

  Read the article

• Linux buffer cache effect on IO writes?

  - by Patrick LeBoutillier

  I'm copying large files (3 x 30G) between 2 filesystems on a Linux server (kernel 2.6.37, 16 cores, 32G RAM) and I'm getting poor performance. I suspect that the usage of the buffer cache is killing the I/O performance. To try and narrow down the problem I used fio directly on the SAS disk to monitor the performance. Here is the output of 2 fio runs (the first with direct=1, the second one direct=0): Config: [test] rw=write blocksize=32k size=20G filename=/dev/sda # direct=1 Run 1: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/205M /s] [0/6K iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4667 write: io=20,480MB, bw=199MB/s, iops=6,381, runt=102698msec clat (usec): min=104, max=13,388, avg=152.06, stdev=72.43 bw (KB/s) : min=192448, max=213824, per=100.01%, avg=204232.82, stdev=4084.67 cpu : usr=3.37%, sys=16.55%, ctx=655410, majf=0, minf=29 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 250=99.50%, 500=0.45%, 750=0.01%, 1000=0.01% lat (msec): 2=0.01%, 4=0.02%, 10=0.01%, 20=0.01% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=199MB/s, minb=204MB/s, maxb=204MB/s, mint=102698msec, maxt=102698msec Disk stats (read/write): sda: ios=0/655238, merge=0/0, ticks=0/79552, in_queue=78640, util=76.55% Run 2: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/0K /s] [0/0 iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4733 write: io=20,480MB, bw=91,265KB/s, iops=2,852, runt=229786msec clat (usec): min=16, max=127K, avg=349.53, stdev=4694.98 bw (KB/s) : min=56013, max=1390016, per=101.47%, avg=92607.31, stdev=167453.17 cpu : usr=0.41%, sys=6.93%, ctx=21128, majf=0, minf=33 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 20=5.53%, 50=93.89%, 100=0.02%, 250=0.01%, 500=0.01% lat (msec): 2=0.01%, 4=0.01%, 10=0.01%, 20=0.01%, 50=0.12% lat (msec): 100=0.38%, 250=0.04% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=91,265KB/s, minb=93,455KB/s, maxb=93,455KB/s, mint=229786msec, maxt=229786msec Disk stats (read/write): sda: ios=8/79811, merge=7/7721388, ticks=9/32418456, in_queue=32471983, util=98.98% I'm not knowledgeable enough with fio to interpret the results, but I don't expect the overall performance using the buffer cache to be 50% less than with O_DIRECT. Can someone help me interpret the fio output? Are there any kernel tunings that could fix/minimize the problem? Thanks a lot,

  Read the article

• Linux buffer cache effect on IO writes?

  - by Patrick LeBoutillier

  Hi, I'm copying large files (3 x 30G) between 2 filesystems on a Linux server (kernel 2.6.37, 16 cores, 32G RAM) and I'm getting poor performance. I suspect that the usage of the buffer cache is killing the I/O performance. To try and narrow down the problem I used fio directly on the SAS disk to monitor the performance. Here is the output of 2 fio runs (the first with direct=1, the second one direct=0): Config: [test] rw=write blocksize=32k size=20G filename=/dev/sda # direct=1 Run 1: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/205M /s] [0/6K iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4667 write: io=20,480MB, bw=199MB/s, iops=6,381, runt=102698msec clat (usec): min=104, max=13,388, avg=152.06, stdev=72.43 bw (KB/s) : min=192448, max=213824, per=100.01%, avg=204232.82, stdev=4084.67 cpu : usr=3.37%, sys=16.55%, ctx=655410, majf=0, minf=29 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 250=99.50%, 500=0.45%, 750=0.01%, 1000=0.01% lat (msec): 2=0.01%, 4=0.02%, 10=0.01%, 20=0.01% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=199MB/s, minb=204MB/s, maxb=204MB/s, mint=102698msec, maxt=102698msec Disk stats (read/write): sda: ios=0/655238, merge=0/0, ticks=0/79552, in_queue=78640, util=76.55% Run 2: test: (g=0): rw=write, bs=32K-32K/32K-32K, ioengine=sync, iodepth=1 Starting 1 process Jobs: 1 (f=1): [W] [100.0% done] [0K/0K /s] [0/0 iops] [eta 00m:00s] test: (groupid=0, jobs=1): err= 0: pid=4733 write: io=20,480MB, bw=91,265KB/s, iops=2,852, runt=229786msec clat (usec): min=16, max=127K, avg=349.53, stdev=4694.98 bw (KB/s) : min=56013, max=1390016, per=101.47%, avg=92607.31, stdev=167453.17 cpu : usr=0.41%, sys=6.93%, ctx=21128, majf=0, minf=33 IO depths : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0% submit : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% complete : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0% issued r/w: total=0/655360, short=0/0 lat (usec): 20=5.53%, 50=93.89%, 100=0.02%, 250=0.01%, 500=0.01% lat (msec): 2=0.01%, 4=0.01%, 10=0.01%, 20=0.01%, 50=0.12% lat (msec): 100=0.38%, 250=0.04% Run status group 0 (all jobs): WRITE: io=20,480MB, aggrb=91,265KB/s, minb=93,455KB/s, maxb=93,455KB/s, mint=229786msec, maxt=229786msec Disk stats (read/write): sda: ios=8/79811, merge=7/7721388, ticks=9/32418456, in_queue=32471983, util=98.98% I'm not knowledgeable enough with fio to interpret the results, but I don't expect the overall performance using the buffer cache to be 50% less than with O_DIRECT. Can someone help me interpret the fio output? Are there any kernel tunings that could fix/minimize the problem? Thanks a lot,

  Read the article

• socket() failed: No buffer space available) while connecting to upstream,

  - by alfish

  On my ubuntu 10.04 VPS, I get regular 500 error on nginx (0.7.??)+ fcgi web server running a durpal site. and when I trace the nginx error log I see plenty of these: socket() failed: No buffer space available) while connecting to upstream ..., I have tried differnt combination configs but none fixed the problem. Currently I have 3 nginx workers, Keep-alive time out 15 seconds and and PHP_FCGI_CHILDREN=5 PHP_FCGI_MAX_REQUESTS=1000 I really appreciate if you Can you suggest a solution to this annoying problem.

  Read the article

• Increase Scrollback Buffer in Putty

  - by Eric J.

  I'm running Putty on Windows and can't find an option to increase the size of the scrollback buffer . It seems set at about 4 pages of text. Am I missing something?

  Read the article

• java.sql.SQLException: ORA-06502: PL/SQL: numeric or value error: character string buffer too small

  - by jack

  Hi I got an email from a user when he sees the following error output when he's using our web site. java.sql.SQLException: ORA-06502: PL/SQL: numeric or value error: character string buffer too small ORA-06512: at "WEB_OWNER.SSFP_GET_WE_OBJ", line 300 ORA-06512: at line 1 at oracle.jdbc.driver.DatabaseError.throwSqlException(DatabaseError.java:137) at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:315) at oracle.jdbc.driver.T4CTTIoer.processError(T4CTTIoer.java:281) This is error from oracle webconnect, Oracle Application Server Containers for J2EE 10g (10.1.2.3.0). Any idea?

  Read the article

• Buffer requests in nginx while a symlink switches on backend

  - by Quintin Par

  In a release deployment I would like to buffer client requests that come to nginx(in reverse proxy) mode to be buffered for possibily 1-2 seconds while a pdsh request is sent to switch symlinks on the back end server to /var/www/html/current . After the switch is complete, I would want to release the buffering while avoiding a herd clash. Is this possible in nginx? Can someone help? Edit: The idea is not to loose requests and from nginx forums I've come to know that retries can sometimes results in CPU spins

  Read the article

• MCE7 how to get video from livetv buffer

  - by Vnuk

  I'm watching something on my MCE and I can jump back to see something. How can I (if at all) extract a piece of that video that is recorded on my drive somewhere? If I press record it starts recording from now and past live buffer is lost. Is there a plugin that can do this?

  Read the article

• Reverse Proxy that does not buffer uploads

  - by tsuraan

  From what I've seen of various reverse proxies (nginx, apache, varnish), they seem to buffer file uploads to disk before handing them off to the service they're proxying for. I need a reverse proxy that doesn't do this; I have a system that handles uploads itself, and buffering uploaded files to disk is not something that works for me. Does anybody know of a proxy server that can be configured to just pass traffic through to the proxied services without doing any buffering to disk?

  Read the article

• Python: convert buffer type of SQLITE column into string

  - by Volatil3

  I am new to Python 2.6. I have been trying to fetch date datetime value which is in yyyy-mm-dd hh:m:ss format back in my Python program. On checking the column type in Python I get the error: 'buffer' object has no attribute 'decode'. I want to use the strptime() function to split the date data and use it but I can't find how to convert a buffer to string. The following is a sample of my code (also available here): conn = sqlite3.connect("mrp.db.db", detect_types=sqlite3.PARSE_DECLTYPES) cursor = conn.cursor() qryT = """ SELECT dateDefinitionTest FROM t WHERE IDproject = 4 AND IDstatus = 5 ORDER BY priority, setDate DESC """ rec = (4,4) cursor.execute(qryT,rec) resultsetTasks = cursor.fetchall() cursor.close() # closing the resultset for item in resultsetTasks: taskDetails = {} _f = item[10].decode("utf-8") The exception I get is: 'buffer' object has no attribute 'decode'

  Read the article

• Visualize the depth buffer

  - by Thanatos

  I'm attempting to visualize the depth buffer for debugging purposes, by drawing it on top of the actual rendering when a key is pressed. It's mostly working, but the resulting image appears to be zoomed in. (It's not just the original image, in an odd grayscale) Why is it not the same size as the color buffer? This is what I'm using the view the depth buffer: void get_gl_size(int &width, int &height) { int iv[4]; glGetIntegerv(GL_VIEWPORT, iv); width = iv[2]; height = iv[3]; } void visualize_depth_buffer() { int width, height; get_gl_size(width, height); float *data = new float[width * height]; glReadPixels(0, 0, width, height, GL_DEPTH_COMPONENT, GL_FLOAT, data); glDrawPixels(width, height, GL_LUMINANCE, GL_FLOAT, data); delete [] data; }

  Read the article

• realloc()ing memory for a buffer used in recv()

  - by Hristo

  I need to recv() data from a socket and store it into a buffer, but I need to make sure get all of the data so I have things in a loop. So to makes sure I don't run out of room in my buffer, I'm trying to use realloc to resize the memory allocated to the buffer. So far I have: // receive response int i = 0; int amntRecvd = 0; char *pageContentBuffer = (char*) malloc(4096 * sizeof(char)); while ((amntRecvd = recv(proxySocketFD, pageContentBuffer + i, 4096, 0)) > 0) { i += amntRecvd; realloc(pageContentBuffer, 4096 + sizeof(pageContentBuffer)); } However, this doesn't seem to be working properly since Valgrind is complaining "valgrind: the 'impossible' happened:". Any advice as to how this should be done properly? Thanks, Hristo

  Read the article

< Previous Page | 6 7 8 9 10 11 12 13 14 15 16 17  | Next Page >