Search Results

Search found 18148 results on 726 pages for 'performance monitor'.

Page 194/726 | < Previous Page | 190 191 192 193 194 195 196 197 198 199 200 201  | Next Page >

• What is wrong with the performance ?

  - by Innovative

  I created an application ( Creating up to 1,00,000 files in a folder ) with 2 different languages ( C#.Net & JAva ) with file concepts and its best methods. The Application that I created with .NET takes just 3 seconds to create all the files. At the same time, with Java, it takes 15 seconds. Why is it so ?

  Read the article

• High performance distributed asynchronous RPC in java

  - by unludo

  I would like to do RPC to a list of clients with the following requirements: the server does not know the clients (implies a kind of broker?) and the cleints do not know the server there may be several clients - they share the load to treat the RPC The RPC is asynchronous very fast (round-trip < 1ms) optional : offers a fail-over mechanism. It can be done with underlying tools which are not really intended for that (Hazelcast is an example). What would you use for such requirements? Thanks!

  Read the article

• C/C++ Pre vs Post decrement and increment operators performance difference

  - by TimK

  In some of my programs I have noticed that in some cases there is a difference in speed when using either the pre(++i or --i) and post(i++ or i--) decrement operators. Sometimes pre is faster and other times post is faster. I can't seem to figure out why. Could someone please explain this to me?

  Read the article

• SQL Server insert performance with and without primary key

  - by Eric

  Summary: I have a table populated via the following: insert into the_table (...) select ... from some_other_table Running the above query with no primary key on the_table is ~15x faster than running it with a primary key, and I don't understand why. The details: I think this is best explained through code examples. I have a table: create table the_table ( a int not null, b smallint not null, c tinyint not null ); If I add a primary key, this insert query is terribly slow: alter table the_table add constraint PK_the_table primary key(a, b); -- Inserting ~880,000 rows insert into the_table (a,b,c) select a,b,c from some_view; Without the primary key, the same insert query is about 15x faster. However, after populating the_table without a primary key, I can add the primary key constraint and that only takes a few seconds. This one really makes no sense to me. More info: The estimated execution plan shows 0% total query time spent on the clustered index insert SQL Server 2008 R2 Developer edition, 10.50.1600 Any ideas?

  Read the article

• Why Swift is 100 times slower than C in this image processing test?

  - by xiaobai

  Like many other developers I have been very excited at the new Swift language from Apple. Apple has boasted its speed is faster than Objective C and can be used to write operating system. And from what I learned so far, it's a very type-safe language and able to have precisely control over the exact data type (like integer length). So it does look like having good potential handling performance critical tasks, like image processing, right? That's what I thought before I carried out a quick test. The result really surprised me. Here is a much simplified image alpha blending code snippet in C: test.c: #include <stdio.h> #include <stdint.h> #include <string.h> uint8_t pixels[640*480]; uint8_t alpha[640*480]; uint8_t blended[640*480]; void blend(uint8_t* px, uint8_t* al, uint8_t* result, int size) { for(int i=0; i<size; i++) { result[i] = (uint8_t)(((uint16_t)px[i]) *al[i] /255); } } int main(void) { memset(pixels, 128, 640*480); memset(alpha, 128, 640*480); memset(blended, 255, 640*480); // Test 10 frames for(int i=0; i<10; i++) { blend(pixels, alpha, blended, 640*480); } return 0; } I compiled it on my Macbook Air 2011 with the following command: gcc -O3 test.c -o test The 10 frame processing time is about 0.01s. In other words, it takes the C code 1ms to process one frame: $ time ./test real 0m0.010s user 0m0.006s sys 0m0.003s Then I have a Swift version of the same code: test.swift: let pixels = UInt8[](count: 640*480, repeatedValue: 128) let alpha = UInt8[](count: 640*480, repeatedValue: 128) let blended = UInt8[](count: 640*480, repeatedValue: 255) func blend(px: UInt8[], al: UInt8[], result: UInt8[], size: Int) { for(var i=0; i<size; i++) { var b = (UInt16)(px[i]) * (UInt16)(al[i]) result[i] = (UInt8)(b/255) } } for i in 0..10 { blend(pixels, alpha, blended, 640*480) } The build command line is: xcrun swift -O3 test.swift -o test Here I use the same O3 level optimization flag to make the comparison hopefully fair. However, the resulting speed is 100 time slower: $ time ./test real 0m1.172s user 0m1.146s sys 0m0.006s In other words, it takes Swift ~120ms to processing one frame which takes C just 1 ms. I also verified the memory initialization time in both test code are very small compared to the blend processing function time. What happened?

  Read the article

• Static constructors cause a performance over head ?

  - by AB Kolan

  Recently read in a article on dotnetpearls.com here saying that static ctors take a substantial amount of perfomance hit. Could not fathom why ?

  Read the article

• Performance implications of using a variable versus a magic number

  - by Nathan

  I'm often confused by this. I've always been taught to name numbers I use often using variables or constants, but if it reduces the efficiency of the program, should I still do it? Heres an example: private int CenterText(Font font, PrintPageEventArgs e, string text) { int recieptCenter = 125; int stringLength = Convert.ToInt32(e.Graphics.MeasureString(text, font)); return recieptCenter - stringLength / 2; } The above code is using named variables, but runs slower then this code: private int CenterText(Font font, PrintPageEventArgs e, string text) { return 125 - Convert.ToInt32(e.Graphics.MeasureString(text, font) / 2); } In this example, the difference in execution time is minimal, but what about in larger blocks of code?

  Read the article

• How to controll (and emulate) monitor/projector on Linux?

  - by klew

  I would like to write some C/C++ program to controll monitor/projector attached to my computer. I've googled around, but I couldn't find almost anything that would lead me in some direction. Do you know where I can learn about it? What I would like to do is to write "Hello world" on an external monitor. It would be also nice to have some "monitor emulator" that will act as physicaly attached monitor (or to write program to emulate it). I'm not sure if it is possible. I couldn't find anything like this. I'm working on Ubuntu 9.10 with Gnome.

  Read the article

• Improve SQL query performance

  - by Anax

  I have three tables where I store actual person data (person), teams (team) and entries (athlete). The schema of the three tables is: In each team there might be two or more athletes. I'm trying to create a query to produce the most frequent pairs, meaning people who play in teams of two. I came up with the following query: SELECT p1.surname, p1.name, p2.surname, p2.name, COUNT(*) AS freq FROM person p1, athlete a1, person p2, athlete a2 WHERE p1.id = a1.person_id AND p2.id = a2.person_id AND a1.team_id = a2.team_id AND a1.team_id IN ( SELECT id FROM team, athlete WHERE team.id = athlete.team_id GROUP BY team.id HAVING COUNT(*) = 2 ) GROUP BY p1.id ORDER BY freq DESC Obviously this is a resource consuming query. Is there a way to improve it?

  Read the article

• Sql Server query performance

  - by Macros

  I have a stored procedure on a busy database which constantly come out top in the list of expensive queries (by some way). The query is very simple, it takes a single parameter (@ID, int) which is the primary key of the table, and selects the record that matches that ID. The primary key is an identity field with a clustered index, so I am stumped as to how to optimise this any further? The query is as follows CREATE PROCEDURE [dbo].[P_Call_Get] @ID int = null AS select ID, AppID, AgentID, AgentLogin, Ext, VDN, VDNName, Skill, SkillName, CallFrom, TelNoFrom, ParentCallID, CallStart, ACWStart, CallEnd, Outcome, StageID, TxTo, TxSuccess, ServiceID, DiallerID, CRC, TSCallID, CallDirection, [Manual], CallBackAgent, CallBackDateTime, Notes from P_Call where (ID = @ID or @ID is null) Not sure the best way to post the execution plan - all it shows is that 100% of the operation is taken up by the clustered index scan

  Read the article

• Performance issue between builds

  - by DeadMG

  I've been developing a small indie game in my spare time and have run across an inexplicable issue. Some builds of the game will randomly run several hundred frames per second slower than other builds. For example, when rendering some text and no 3D scene, I can achieve 1800FPS on my own hardware. Add one 3D sphere (10k verts, pixel shaded), achieve 1700 FPS. Add two more spheres, achieve 800 FPS. Remove all spheres, achieve 1100FPS- even though the code now renders the same scene as I previously achieved at 1800FPS, which is just the FPS counter being rendered. I've tried rebuilding and cleaning the project and rebooting the compiler. This is in Release mode and I turned on all the optimizations I could find. Any suggestions as to the cause? I ran a quick profile, and Visual Studio seems to think that over 90% of my time was spent in D3D9_43.dll, suggesting that it's not a bug in my app, which doesn't explain why it manifests in only some builds. I rebooted my machine and it's back up to 1800FPS. I think it's a bug in the DirectX SDK tools (amongst many others). Going to delete this question.

  Read the article

• Eclipse plugin to measure programmer performance/stats

  - by trenki

  Does anyone know of an Eclipse plugin that can give me some stats about my behavior/usage of the Eclipse IDE? There are quite a few things I would like to know: How often/when do I invoke the "Build All" command (through Ctrl+B) How often does compilation fail/succeed (+ number of errors/warnings) How often do I hit Backspace? (I do that way to often; If pressing that key would give a nasty sound I would in time learn to type correctly in the first place) How many characters/lines of code that I typed do I delete (possibly quite immediately) How (effective/efficient/...) is my Mouse/Keyboard/IDE usage? (Kinda like measuring APM in StarCraft; this could be fun) If there is no such Eclipse plugin around, how complex and time consuming would It be to write a plugin that can accomplish the above?

  Read the article

• Database with great read performance

  - by mscava

  I have 10 tables from which 4 contain each up to million rows. All values are inserted at once, and afterwards I only read the data many times. I am searching for a database that would perform greatly when it comes to selecting, joining or other reading etc. What is the most recommended option?

  Read the article

• MySQL TEXT field performance

  - by Jonathon

  I have several TEXT and/or MEDIUMTEXT fields in each of our 1000 MySQL tables. I now know that TEXT fields are written to disk rather than in memory when queried. Is that also true even if that field is not called in the query? For example, if I have a table (tbExam) with 2 fields (id int(11) and comment text) and I run SELECT id FROM tbExam, does MySQL still have to write that to disk before returning results or will it run that query in memory? I am trying to figure out if I need to reconfigure our actual db tables to switch to varchar(xxxx) or keep the text fields and reconfigure the queries.

  Read the article

• Are UIView's in TableViewCells cause for performance problems?

  - by iFloh

  Hi, I have a UITableView of UITableViewCells that so far performed well. I now aded a simple UIView to my UITableViewCell as a color indicator (backgroundColor only, no other content or attributes). Running the tableview again I realise scrolling is not as smooth as it used to be. The tableView suddenly also takes long time (~1+ secs) to come back again when returning from a subview (e.g. return from the view that gets pushed upon selection of a cell). Any ideas why the added UIView would be problematic? Why are the labels (being a subclass of UIView) not cause for similar issues?

  Read the article

• Why fill() and copy() of Collections in java is implemented this way

  - by Priyank Doshi

  According to javadoc... Collections.fill() is written as below : public static <T> void fill(List<? super T> list, T obj) { int size = list.size(); if (size < FILL_THRESHOLD || list instanceof RandomAccess) { for (int i=0; i<size; i++) list.set(i, obj); } else { ListIterator<? super T> itr = list.listIterator(); for (int i=0; i<size; i++) { itr.next(); itr.set(obj); } } } Its easy to understand why they didn't use listIterator for if (size < FILL_THRESHOLD || list instanceof RandomAccess) condition as of RandomAccess. But whats the use of size < FILL_THRESHOLD in above? I mean is there any significant performance benefit over using iterator for size>=FILL_THRESHOLD and not for size < FILL_THRESHOLD ? I see the same approach for Collections.copy() also : public static <T> void copy(List<? super T> dest, List<? extends T> src) { int srcSize = src.size(); if (srcSize > dest.size()) throw new IndexOutOfBoundsException("Source does not fit in dest"); if (srcSize < COPY_THRESHOLD || (src instanceof RandomAccess && dest instanceof RandomAccess)) { for (int i=0; i<srcSize; i++) dest.set(i, src.get(i)); } else { ListIterator<? super T> di=dest.listIterator(); ListIterator<? extends T> si=src.listIterator(); for (int i=0; i<srcSize; i++) { di.next(); di.set(si.next()); } } } FYI: private static final int FILL_THRESHOLD = 25; private static final int COPY_THRESHOLD = 10;

  Read the article

• SQL Server dilemma, performance

  - by Woland

  Hello I am creating app where user can save options witch one is better? to save into user table varchar feeld smthing like ('1,23,4354,34,3') query for this is select * from data where CHARINDEX ( 'L', Providers , 0 ) 0 create other table where user options are and just add rows select * from data where Providers in (select Providers from userdata where userid=100) thanks for help

  Read the article

• Improve C function performance with cache locality?

  - by Christoper Hans

  I have to find a diagonal difference in a matrix represented as 2d array and the function prototype is int diagonal_diff(int x[512][512]) I have to use a 2d array, and the data is 512x512. This is tested on a SPARC machine: my current timing is 6ms but I need to be under 2ms. Sample data: [3][4][5][9] [2][8][9][4] [6][9][7][3] [5][8][8][2] The difference is: |4-2| + |5-6| + |9-5| + |9-9| + |4-8| + |3-8| = 2 + 1 + 4 + 0 + 4 + 5 = 16 In order to do that, I use the following algorithm: int i,j,result=0; for(i=0; i<4; i++) for(j=0; j<4; j++) result+=abs(array[i][j]-[j][i]); return result; But this algorithm keeps accessing the column, row, column, row, etc which make inefficient use of cache. Is there a way to improve my function?

  Read the article

• Hibernate Performance Best Practice?

  - by user829237

  Im writing a Web application using Hibernate 3. So, after a while i noticed that something was slow. So i tested hibernate profiler and found that hibernate will make unreasonably many db-calls for simple operation. The reason is ofcourse that i load an Object (this object has several "parents") and these "parents" have other "parents". So basicly hibernate loads them all, even though i just need the basic object. Ok, so i looked into lazy-loading. Which lead me into the Lazyloading-exception, because i have a MVC webapp. So now i'm a bit confused as to what is my best approach to this. Basicly all I need is to update a single field on an object. I already have the object-key. Should I: 1. Dig into Lazy-loading. And then rewrite my app for a open-session-view? 2. Dig into lazy-loading. And then rewrite my dao's to be more specific. E.g. writing DAO-methods that will return objects instanciated with only whats necessary for each use-case? Could be a lot of extra methods... 3. Scratch hibernate and do it myself? 4. Cant really think of other solutions right now. Any suggestions? What is the best practice?

  Read the article

• High performance SSL client with connection pooling session caching etch on Windows

  - by Yucong Sun

  Hi there, I'm looking for a fast SSL socket client basically talks SSL with remote service and I need features like connection pooling/limiting so that my other client calling this ssl socket can basically issue as many as requests and it will handle it peacefully. I looked into openssl s_client, but it says it's only for testing purpose, is there a handy tool avaiable? STunnel seems to be okay, but don't have that connection limiting option I need. Cheers.

  Read the article

• Javascript Library Performance

  - by Steve

  Hi..I recently bumped into this link http://mootools.net/slickspeed/ I was amazed to see that the prototype is extremely slow. So I would like to know your comments on this?

  Read the article

• Database Design and Performance : mySQL and PHP

  - by jasmine

  I have made an image for this question: What is the right design? All site's block inserting in only one table or separate tables? Thanks in advance

  Read the article

• How do you run XBMC on nvidia dual screen and stop it from taking over the keyboard and mouse?

  - by Paul Swartout

  I have set up dual screen under Ubuntu 12.04. I have a GeForce 8500 GT and have used the nVidia control panel to set up dual screen in "Separate screen mode". Here's the resulting xorg.conf # nvidia-settings: X configuration file generated by nvidia-settings # nvidia-settings: version 295.33 (buildd@zirconium) Fri Mar 30 13:38:49 UTC 2012 Section "ServerLayout" Identifier "Layout0" Screen 0 "Screen0" 0 0 Screen 1 "Screen1" RightOf "Screen0" InputDevice "Keyboard0" "CoreKeyboard" InputDevice "Mouse0" "CorePointer" Option "Xinerama" "0" EndSection Section "Files" EndSection Section "InputDevice" # generated from default Identifier "Mouse0" Driver "mouse" Option "Protocol" "auto" Option "Device" "/dev/psaux" Option "Emulate3Buttons" "no" Option "ZAxisMapping" "4 5" EndSection Section "InputDevice" # generated from default Identifier "Keyboard0" Driver "kbd" EndSection Section "Monitor" # HorizSync source: edid, VertRefresh source: edid Identifier "Monitor0" VendorName "Unknown" ModelName "Maxdata/Belinea B1925S1W" HorizSync 31.0 - 83.0 VertRefresh 56.0 - 75.0 Option "DPMS" EndSection Section "Monitor" # HorizSync source: builtin, VertRefresh source: builtin Identifier "Monitor1" VendorName "Unknown" ModelName "CRT-1" HorizSync 28.0 - 55.0 VertRefresh 43.0 - 72.0 Option "DPMS" EndSection Section "Device" Identifier "Device0" Driver "nvidia" VendorName "NVIDIA Corporation" BoardName "GeForce 8500 GT" BusID "PCI:1:0:0" Screen 0 EndSection Section "Device" Identifier "Device1" Driver "nvidia" VendorName "NVIDIA Corporation" BoardName "GeForce 8500 GT" BusID "PCI:1:0:0" Screen 1 EndSection Section "Screen" Identifier "Screen0" Device "Device0" Monitor "Monitor0" DefaultDepth 24 Option "TwinView" "0" Option "metamodes" "CRT-0: nvidia-auto-select +0+0" SubSection "Display" Depth 24 EndSubSection EndSection Section "Screen" # Removed Option "metamodes" "CRT-1: 1280x768 +0+0" Identifier "Screen1" Device "Device1" Monitor "Monitor1" DefaultDepth 24 Option "TwinView" "0" Option "metamodes" "CRT-1: 1360x768_60 +0+0" SubSection "Display" Depth 24 EndSubSection EndSection All well and good and I have a nice blank XWindow displayed on my TV (the 2nd monitor). I then fire up XBMC from a terminal on the PC monitor using DISPLAY=:0.1 xbmc XBMC fires up quite nicely on the TV however I can no longer use the main PC monitor / mouse / keyboard as XBMC on the TV screen seems to have focus. I was hoping to have XBMC running on the TV and let the kids use the MCE remote whilst I get on with my work on the PC monitor. Does anyone have any idea how to overcome this? I'm presuming there's some xorg.conf fun and games needed but I've no idea where to start to be honest.

  Read the article

• Ubuntu 13.10. After login, no desktop displayed. Two Nvidia Graphics Cards, Four Monitors

  - by jmerkow

  I am working on an issue with my Ubuntu 13.10 installation. I am attempting to get 4 monitors up and running but I am having some trouble. So far, I installed and updated to the latest NVIDIA drivers (331.20). Initially X would not start (after installation) so I replaced my xorg.conf with xorg.conf.failsafe. This fixed that problem, but then I tried to enable the other 2 monitors (other video card) and xorg fails to start once again (after I login there is no desktop). I am fairly new to linux but I am not a complete beginner, but I'm not comfortable poking around too much on my own to troubleshoot yet.... lspci -nn | grep VGA: 03:00.0 VGA compatible controller [0300]: NVIDIA Corporation GF110 [GeForce GTX 570 Rev. 2] [10de:1086] (rev a1) 05:00.0 VGA compatible controller [0300]: NVIDIA Corporation GF110 [GeForce GTX 580] [10de:1080] (rev a1) It seems that the nvidia-settings tool does not result in a good xorg.conf file. Here it is: # nvidia-settings: X configuration file generated by nvidia-settings # nvidia-settings: version 331.20 (buildmeister@swio-display-x86-rhel47-05) Wed Oct 30 18:20:32 PDT 2013 Section "ServerLayout" Identifier "Default Layout" Screen 0 "Screen0" 0 0 Screen 1 "Screen1" RightOf "Screen0" InputDevice "Keyboard0" "CoreKeyboard" InputDevice "Mouse0" "CorePointer" Option "Xinerama" "1" EndSection ... Section "Monitor" Identifier "Configured Monitor" EndSection Section "Monitor" Identifier "Monitor0" VendorName "Unknown" ModelName "SHARP HDMI" HorizSync 15.0 - 68.0 VertRefresh 55.0 - 76.0 EndSection Section "Monitor" Identifier "Monitor1" VendorName "Unknown" ModelName "Samsung SyncMaster" HorizSync 0.0 - 0.0 VertRefresh 0.0 EndSection Section "Device" Identifier "Configured Video Device" Driver "vesa" EndSection Section "Device" Identifier "Device0" Driver "nvidia" VendorName "NVIDIA Corporation" BoardName "GeForce GTX 570" BusID "PCI:3:0:0" EndSection Section "Device" Identifier "Device1" Driver "nvidia" VendorName "NVIDIA Corporation" BoardName "GeForce GTX 580" BusID "PCI:5:0:0" EndSection Section "Screen" Identifier "Default Screen" Device "Configured Video Device" Monitor "Configured Monitor" EndSection Section "Screen" Identifier "Screen0" Device "Device0" Monitor "Monitor0" DefaultDepth 24 Option "Stereo" "0" Option "nvidiaXineramaInfoOrder" "DFP-1" Option "metamodes" "HDMI-0: nvidia-auto-select +640+0, DVI-I-3: nvidia-auto-select +0+1080" Option "SLI" "Off" Option "MultiGPU" "Off" Option "BaseMosaic" "off" SubSection "Display" Depth 24 EndSubSection EndSection Section "Screen" Identifier "Screen1" Device "Device1" Monitor "Monitor1" DefaultDepth 24 Option "Stereo" "0" Option "metamodes" "DVI-I-2: nvidia-auto-select +0+0" Option "SLI" "Off" Option "MultiGPU" "Off" Option "BaseMosaic" "off" SubSection "Display" Depth 24 EndSubSection EndSection Section "Extensions" Option "Composite" "Disable" EndSection

  Read the article

• Of these 3 methods for reading linked lists from shared memory, why is the 3rd fastest?

  - by Joseph Garvin

  I have a 'server' program that updates many linked lists in shared memory in response to external events. I want client programs to notice an update on any of the lists as quickly as possible (lowest latency). The server marks a linked list's node's state_ as FILLED once its data is filled in and its next pointer has been set to a valid location. Until then, its state_ is NOT_FILLED_YET. I am using memory barriers to make sure that clients don't see the state_ as FILLED before the data within is actually ready (and it seems to work, I never see corrupt data). Also, state_ is volatile to be sure the compiler doesn't lift the client's checking of it out of loops. Keeping the server code exactly the same, I've come up with 3 different methods for the client to scan the linked lists for changes. The question is: Why is the 3rd method fastest? Method 1: Round robin over all the linked lists (called 'channels') continuously, looking to see if any nodes have changed to 'FILLED': void method_one() { std::vector<Data*> channel_cursors; for(ChannelList::iterator i = channel_list.begin(); i != channel_list.end(); ++i) { Data* current_item = static_cast<Data*>(i->get(segment)->tail_.get(segment)); channel_cursors.push_back(current_item); } while(true) { for(std::size_t i = 0; i < channel_list.size(); ++i) { Data* current_item = channel_cursors[i]; ACQUIRE_MEMORY_BARRIER; if(current_item->state_ == NOT_FILLED_YET) { continue; } log_latency(current_item->tv_sec_, current_item->tv_usec_); channel_cursors[i] = static_cast<Data*>(current_item->next_.get(segment)); } } } Method 1 gave very low latency when then number of channels was small. But when the number of channels grew (250K+) it became very slow because of looping over all the channels. So I tried... Method 2: Give each linked list an ID. Keep a separate 'update list' to the side. Every time one of the linked lists is updated, push its ID on to the update list. Now we just need to monitor the single update list, and check the IDs we get from it. void method_two() { std::vector<Data*> channel_cursors; for(ChannelList::iterator i = channel_list.begin(); i != channel_list.end(); ++i) { Data* current_item = static_cast<Data*>(i->get(segment)->tail_.get(segment)); channel_cursors.push_back(current_item); } UpdateID* update_cursor = static_cast<UpdateID*>(update_channel.tail_.get(segment)); while(true) { if(update_cursor->state_ == NOT_FILLED_YET) { continue; } ::uint32_t update_id = update_cursor->list_id_; Data* current_item = channel_cursors[update_id]; if(current_item->state_ == NOT_FILLED_YET) { std::cerr << "This should never print." << std::endl; // it doesn't continue; } log_latency(current_item->tv_sec_, current_item->tv_usec_); channel_cursors[update_id] = static_cast<Data*>(current_item->next_.get(segment)); update_cursor = static_cast<UpdateID*>(update_cursor->next_.get(segment)); } } Method 2 gave TERRIBLE latency. Whereas Method 1 might give under 10us latency, Method 2 would inexplicably often given 8ms latency! Using gettimeofday it appears that the change in update_cursor-state_ was very slow to propogate from the server's view to the client's (I'm on a multicore box, so I assume the delay is due to cache). So I tried a hybrid approach... Method 3: Keep the update list. But loop over all the channels continuously, and within each iteration check if the update list has updated. If it has, go with the number pushed onto it. If it hasn't, check the channel we've currently iterated to. void method_three() { std::vector<Data*> channel_cursors; for(ChannelList::iterator i = channel_list.begin(); i != channel_list.end(); ++i) { Data* current_item = static_cast<Data*>(i->get(segment)->tail_.get(segment)); channel_cursors.push_back(current_item); } UpdateID* update_cursor = static_cast<UpdateID*>(update_channel.tail_.get(segment)); while(true) { for(std::size_t i = 0; i < channel_list.size(); ++i) { std::size_t idx = i; ACQUIRE_MEMORY_BARRIER; if(update_cursor->state_ != NOT_FILLED_YET) { //std::cerr << "Found via update" << std::endl; i--; idx = update_cursor->list_id_; update_cursor = static_cast<UpdateID*>(update_cursor->next_.get(segment)); } Data* current_item = channel_cursors[idx]; ACQUIRE_MEMORY_BARRIER; if(current_item->state_ == NOT_FILLED_YET) { continue; } found_an_update = true; log_latency(current_item->tv_sec_, current_item->tv_usec_); channel_cursors[idx] = static_cast<Data*>(current_item->next_.get(segment)); } } } The latency of this method was as good as Method 1, but scaled to large numbers of channels. The problem is, I have no clue why. Just to throw a wrench in things: if I uncomment the 'found via update' part, it prints between EVERY LATENCY LOG MESSAGE. Which means things are only ever found on the update list! So I don't understand how this method can be faster than method 2. The full, compilable code (requires GCC and boost-1.41) that generates random strings as test data is at: http://pastebin.com/e3HuL0nr

  Read the article

< Previous Page | 190 191 192 193 194 195 196 197 198 199 200 201  | Next Page >