Search Results

Search found 4335 results on 174 pages for 'wait'.

Page 122/174 | < Previous Page | 118 119 120 121 122 123 124 125 126 127 128 129  | Next Page >

  • Static IP addressing issue in Ubuntu on BeagleBoneBlack Rev C

    - by Stringfellow
    I have my BBB configured to use a static IP address using the following in the file /etc/network/interfaces: allow-hotplug eth0 iface eth0 inet static address 192.168.0.1 netmask 255.255.255.0 network 192.168.0.0 This seems to work ok on boot, but when the ethernet cable is unplugged and then plugged back in, I lose the IP address. Any ideas what's going on here? Another weird symptom: If I boot the BBB with the network cable unplugged, but the switch it's plugged into off, I'll get my static IP. But, when I turn the switch on, I'll get a DHCP-assigned address. This is even though I have it configured with a static IP address. One last thing. If I ifdown etho, the interface will be gone when I do an ifconfig. If I wait a few seconds, though, and then re-run ifconfig, it will reappear, without an IP address. (Before I disabled IPv6, I used to get a IPv4 DHCP address in this case... weird). When that happens, I get a message like this in /var/log/messages: Apr 23 20:32:06 beaglebone kernel: [ 737.170172] libphy: 4a101000.mdio:00 - Link is Up - 100/Full Apr 23 20:32:06 beaglebone kernel: [ 737.170304] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready Here's my uname -a: root@beaglebone:/etc# uname -a Linux beaglebone 3.8.13-bone47 #1 SMP Fri Apr 11 01:36:09 UTC 2014 armv7l GNU/Linux Any ideas what's going on here?

    Read the article

  • Internet compression proxy for low speed broadband?

    - by user23150
    I live in a rural location, using high-latency wireless off a local ISP's tower. My speed tests vary day to day, but I can get around 1Mb up/down. The problem is, I work with large files, uploading and downloading (HD videos, development software, etc.). It can be painful to wait sometimes. Plus I do some side contract game development, and it can be very difficult to playtest with other developers (200ms ping is a good day for me). Now, obviously it's not going to be easy to solve the latency problem without different wireless hardware. But speedwise, I am wondering if I can use some kind of compression technology on a proxy. For instance, my work computer has full access to a 26Mb down, 10Mb up connection, that is totally unused at night and the weekends. If I could run some kind of compression technology on our server, and use it as a proxy to route to my home computer, I could stand to gain some major speed. I realize that by bogging down a system with compression, I could potentially lose whatever speed gain I had. But the proxy server is a quad core xeon, and the receiving computer is a pretty decent i7 computer, so that shouldn't be a concern. I found http://toonel.net/ but it seems more geared toward very slow narrowband users, like dial-up. Plus, I would prefer to just be able to point my browser to a proxy server, rather then install software on my client machine. EDIT I thought about my question a little more, and realize I am going to need to install software on my client in order to decompress, and possible compress (for uploading). That's not a huge deal.

    Read the article

  • Advice on cloning disk

    - by hks
    I'm going to buy a second disk for backup, the same size as my laptops. I want to mount it in a casing via usb and backup an entire hdd every soemtime. That's because I want the posibility to just switch drives in case of something goes wrong. I'm using Linux and obviously the right tool seems to be dd. The thing is that my laptop drive has a speed of around 50-70 MB/s and usb 2.0 is 57 MB/s. So to copy my 250GB disk should take me more than 1 hour if I'm lucky. I can't wait this much. I want some differential backup. I read one of JWZ articles. In it he gives more details for using rsync on Mac. He writes that there is possibility of making rsync'ed disk bootable. So my question is: how to make rsync'ed hdd bootable under Linux or are there other 'quick backup' tools for Linux that would allow me to just swap drives? Or should I just stick to dd :( ?

    Read the article

  • My Hard Drive isn't Working

    - by MeCB
    I never use Safely Remove Hardware with Windows XP. It has been working for me for years with my SD card, mouse, hard drive and memory stick. My hard drive has a USB cable and power cord so I can hook it up to any desktop hard drive. Now my hard drives don't like this and I did not know this till now. I am always careful and to wait till it is all finish accessing the USB before I unplug it. Now three of my hard drives can't be seen by Windows, though the others still work. when I hook it up to another computer it works fine. I use the same USB cable to hook up all of my hard drives one at a time. So my USB cable is good. I think that when I unplugged the hard drive this one time, it had a file it still wanted to see and now only this drive does not work only this computer. Then the same thing happens to my other two hard drive after I used it for a week with the same cables. How can I fix this?

    Read the article

  • How to delete removed devices from a mdadm RAID1?

    - by Kabuto
    I had to replace two hard drives in my RAID1. After adding the two new partitions the old ones are still showing up as removed while the new ones are only added as spare. I've had no luck removing the partitions marked as removed. Here's the RAID in question. Note the two devices (0 and 1) with state removed. $ mdadm --detail /dev/md1 mdadm: metadata format 00.90 unknown, ignored. mdadm: metadata format 00.90 unknown, ignored. /dev/md1: Version : 00.90 Creation Time : Thu May 20 12:32:25 2010 Raid Level : raid1 Array Size : 1454645504 (1387.26 GiB 1489.56 GB) Used Dev Size : 1454645504 (1387.26 GiB 1489.56 GB) Raid Devices : 3 Total Devices : 3 Preferred Minor : 1 Persistence : Superblock is persistent Update Time : Tue Nov 12 21:30:39 2013 State : clean, degraded Active Devices : 1 Working Devices : 3 Failed Devices : 0 Spare Devices : 2 UUID : 10d7d9be:a8a50b8e:788182fa:2238f1e4 Events : 0.8717546 Number Major Minor RaidDevice State 0 0 0 0 removed 1 0 0 1 removed 2 8 34 2 active sync /dev/sdc2 3 8 18 - spare /dev/sdb2 4 8 2 - spare /dev/sda2 How do I get rid of these devices and add the new partitions as active RAID devices? Update I seem to have gotten rid of them. My RAID is resyncing, but the two drives are still marked as spares and are number 3 and 4, which looks wrong. I'll have to wait for the resync to finish. All I did was to fix the metadata error by editing my mdadm.conf and rebooting. I tried rebooting before, but this time it worked for whatever reason. Number Major Minor RaidDevice State 3 8 2 0 spare rebuilding /dev/sda2 4 8 18 1 spare rebuilding /dev/sdb2 2 8 34 2 active sync /dev/sdc2

    Read the article

  • Windows 7 misses keystrokes from internal keyboard after hibernation (on Acer Aspire 5820)

    - by ron
    I face a very strange symptom on my Acer Aspire laptop (with the factory default Win7 install and divers. Windows update running). After waking the computer from hibernation, it is a pain to type, since on average 5-10 keypresses are missing per 100 presses, using the laptop's keyboard. Steps to reproduce: 1) Power off 2) Power on, wait for system to become usable 3) Open notepad, for 5 times do hit 10x the same character. This gives a similar pattern of 50 chars total: xxxxxxxxxxyyyyyyyyyyaaaaaaaaaassssssssssdddddddddd 4) Optionally repeat. Everything is fine this far. 5) Hibernate. 6) Power on and resume. 7) Repeat steps 3)-4). This time approximately 3-5 character will be missing from each 50 characters. What I ruled out: putting to Sleep or just Locking and resuming from there does not cause problem battery / AC usage does not matter net connection does not matter running processes seem to be the same before and after hibernation key press speed doesn't really matter. For the test I use a nominal 3-5 strokes/second beat. plugging in an external USB keyboard works fine, but the built-in one still misbehaves What could be the problem? How could I diagnose if the keypresses arrive in, but get swallowed at some point? (maybe some nasty keyboard handler hook misbehaves?). Update: It seems that pushing the PowerSmart button and toggling to power saving state fixes the problem. Also, toggling it again back to the original state keeps it fixed. So this may be a fine workaround, but is not a conforming solution.

    Read the article

  • Is there a simple context-menu add-in that could make-up for the Windows-7 status bar deficiency?

    - by DanO
    Edit: I initially asked about free disk space and selected item size. It has since been pointed out that the selected item size summary is still availiable natively in the details pane. I had read elsewhere (wikipedia) that this was removed along with disk free space, which is not the case. Only free disk space has been completely removed. Selection size is still availiable. Is there a context menu add-in out there that could show the free disk space of the relevant drive, when you right click? This would go a long way to compensating for one of the only steps backward I’ve discovered in Windows 7 so far. I doubt anyone had created one specifially for this need before windows 7 because this information was previously easily accessible in the status bar. I thought about creating one, but it has been a while since I have messed with the Shell API, and I know there are coders out there who could do it faster and better. If you’ve heard of one, or know of something else to make-up for this Microsoft misstep, I’d appreciate hearing about it. If MS were listing to the community they would already have a powertoy or add-in of some kind to un-break this. (they could release it unsupported even), as there seem to be many power users that are extremely annoyed by this feature removal decision. If anyone has seen something, please post it here. As it has been only 4 days since official Windows 7 release, I'll wait at least a week to chose an answer. Here's a picture of protoype screenshot: SU question 19232 is related.

    Read the article

  • Windows 7 misses keystrokes from internal keyboard after hibernation (on Acer Aspire 5820)

    - by ron
    I face a very strange symptom on my Acer Aspire laptop (with the factory default Win7 install and divers. Windows update running). After waking the computer from hibernation, it is a pain to type, since on average 5-10 keypresses are missing per 100 presses, using the laptop's keyboard. Steps to reproduce: 1) Power off 2) Power on, wait for system to become usable 3) Open notepad, for 5 times do hit 10x the same character. This gives a similar pattern of 50 chars total: xxxxxxxxxxyyyyyyyyyyaaaaaaaaaassssssssssdddddddddd 4) Optionally repeat. Everything is fine this far. 5) Hibernate. 6) Power on and resume. 7) Repeat steps 3)-4). This time approximately 3-5 character will be missing from each 50 characters. What I ruled out: putting to Sleep or just Locking and resuming from there does not cause problem battery / AC usage does not matter net connection does not matter running processes seem to be the same before and after hibernation key press speed doesn't really matter. For the test I use a nominal 3-5 strokes/second beat. plugging in an external USB keyboard works fine, but the built-in one still misbehaves What could be the problem? How could I diagnose if the keypresses arrive in, but get swallowed at some point? (maybe some nasty keyboard handler hook misbehaves?). Update: It seems that pushing the PowerSmart button and toggling to power saving state fixes the problem. Also, toggling it again back to the original state keeps it fixed. So this may be a fine workaround, but is not a conforming solution.

    Read the article

  • Amazon EC2 - Free memory

    - by Damo
    We have an amazon ec2 small instance running and over the past few days we noticed that the memory is going down and down. On the small instance, we are running apache and tomcat6 Tomcat is started with the following JVM parameters -Xms32m -Xmx128m -XX:PermSize=128m -XX:MaxPermSize=256m We use nagios to monitor stuff like updates to apply, free disk space and memory. Everything else is behaving as expected but our memory is going down all the time. Our app receives approx half a million hits a day When I shutdown apache and tomcat, and ran free -m, we had only 594mb of memory free out out of the 1.7gb of memory. Not much else is running on the small instance and when running the top command I cannot see where the memory is going. The app we run on tomcat is a grails webapp. Could there be a possibility that there is a memory leak within our application? I read online and folks say that a small amazon instance is perfect for running apach and tomcat. I found a few posts online that showed how to setup apache and tomcat to limit the memory usage and I have already performed those steps. The memory is not being used up as quick but the memory is still decreasing over time. We have other amazone ec2 small instances running grails apps and the memory is fairly standard on those nodes. But they would not be receiving as much traffic Just to add, when I run the top command on the problem server, I cannot see where all the memory is being used Any help with this is greatly appreciated The output of free -m when run on my server is as follows total used free shared buffers cached Mem: 1657 1380 277 0 158 773 -/+ buffers/cache: 447 1209 Swap: 895 0 895 In your opinion, does this look ok? At what stage would the OS give back memory, would it wait to the memory reaches 0% or is this OS dependent?

    Read the article

  • buagent process has been consuming 100% cpu for two days

    - by Maysam
    The buagent process has been using 100% of cpu since two days ago. I want to terminate this process but I don't know if it's something dangerous or not (I am not much advanced in working with linux, indeed I am very beginner). The only thing that I know is that this process is probably restoring some files. But I think it is not normal for that to take more than two days. Now, do you think it would be OK if I kill this process? What command could I use to do that? I appreciate any help :) p.s. We are hosting a few web sites there. This server is also our Name Server and Mail Server as well. A couple of months a go we had a problem with the server which made us to take a full-backup of all files and then reinstall linux. Yesterday, I selected one of the directories on the backup server and restored that directory to a tmp directory on our linux server. After that, I couldn't restore any other directory because every time I want to do that, it says that there is another restore job running and I have to wait for that. When I use the "top" command I can see that the buagent process is consuming 100% of cpu. So I guess that is the problem. I don't know why it has been taking too long to execute.

    Read the article

  • Network outage caused by SMC8013WG Cable Modem/Router ?

    - by mkocubinski
    At work, we have a basic Class C Network. The gateway/router is a SMC8013WG (stock comcast commercial cable modem), and simple unmanaged switch (HP Pro Curve 1400 24G). The SMC8013WG is our default gateway as well as DHCP server. Periodically, I'd say almost every other day.. the entire network will just stop responding. I won't be able to ping/see the gateway, any computers on our local network, or anything on the internet. The only way to fix this is to unplug the Comcast cable modem, wait, and plug back in. This unfailingly fixes the problem. But this doesn't make much sense to me.. shouldn't the network still be fine locally, since everything is plugged into the switch anyway? Why would resetting the router fix this? Can anyone suggest anything to check to in order to narrow this problem down? Just to be clear.. here is the basic topology: { Internet } -- (12.345.67.89) Comcast Cable Modem (192.168.1.1) -- Switch -- 192.168.1.2-254 P.S. Our IT guy is in about 3 hours a day every other week or so, so.. we're kind of on our own most of the time.

    Read the article

  • Cancelling Window 7 shutdown diables power button

    - by Jens
    Normally, pressing the power button once initiates a shut-down in Windows 7. If any programs are still running that will not quit (e.g. waiting for a dialogue response), Windows overlays the screen with a dialogue allowing the user to cancel the shut-down. I've just noticed that on two different systems here, using this cancel option disables the shut-down via power button. The power button can still be used to kill the system by holding it for a few seconds, using the Start menu button to shut the PC down still works as well. Steps to reproduce: Open Notepad, type a few characters. Do not save. Press the computer's power button. Wait until the dark screen appears. Press cancel. Press the power button again. Notice how nothing happens. What is the reason for this behaviour, and can it be disabled to always try and shut down the PC when the power button is pressed?

    Read the article

  • vi and emacs: comparison? (not flamebait!)

    - by jared
    So, I've been enjoying learning and using vi for the last couple of years. The beauty of vi, for me, is that its UI is a language of movement and action with a very uniform, simple grammar, and which is terse enough that the requisite memorization pays ample dividends in how much more I enjoy working with text (by avoiding boring repetition and eliminating micro-hassles, like that half-second annoying wait while you scroll down the screen). (Note--I don't claim to have expert knowledge of vi, but I get around decently well: comfortable with limited '@' macros and regexp search-and-replace within files; frequently use multiple buffers, tabs, and windows; get around pretty well in the file browser; understand the grammar of actions + movement + subject (as described so aptly in this beautiful SO answer); and had some pretty sweet debugger and ctags integration going with PHP.) I wonder if some emacs folks could take a swing at explaining what emacs does brilliantly, or sum its strengths up in a phrase or two. Spare me the talk about productivity; I'm more interested in conceptual clarity. Lisp-centric answers are okay; I'm learning Scheme on the weekends, and would pick up emacs for that alone (have been using Racket).

    Read the article

  • Windows 7 delayed file delete

    - by GregoryM
    I'm stuck with a pretty rare problem that happens on Windows 7 OS only. Every time I'm deleting the file with *.exe extension through explorer, the file doesn't get deleted immediately. I'm forced to wait for around one-two minutes before the system will delete the file. The main problem is that I cannot develop in such situation, because every time I build my solution, the old executable gets 'deleted', but is still there. So the new one cannot be created by Visual Studio. This problem breaks the Steam update progress and a few other installers functionality too. Fresh installed Win7 doesn't have this kind of trouble, so I guess this must be some bad registry entries or some services. Browsing the internet for solutions I found only this: http://www.sevenforums.com/software/72091-several-minute-delay-when-deleting-any-exe-file.html. But the solution the author found is not working (change the userName :)). Is there any ideas how to find what causes this to happen? BTW: when I place the file into Trash bin, no delay occurs. When I delete file with Total Commander - no delay too. Tech details: Windows 7 x64 Ultimate.

    Read the article

  • Remote Desktop svchost (networkservice) & lsa.exe high cpu usage, hangs on welcome screen

    - by Rohan1
    We have deployed an RDS Farm with 12 virtual RDS servers using Hyper V. Currently some users are not able to log on. After passing credentials to the connection broker, the session hangs on the "Welcome" screen. Using resource monitor we've seen that svchost (with the "networkservice" service) has a CPU usage of 50%, when viewing the wait chain on the process it displays that it's waiting for a lsa.exe to finish. We can't kill any of the users processes, even when trying with taskkill /f. Suspending lsa.exe did work but didn't have any effect. The networkservice also couldn't be restarted. Also, if this happens, the current users logged on to the RDS server can't be displayed. Task manager crashes when viewing the users, RDS service manager crashes when viewing the users (even remotely) and the cmd command "query session" doesn't work. No antivirus is installed on the RDS server. The only thing we can do is rebooting the server, which is not an option because of the fact that other users are in active sessions. Does anyone have ANY idea what's going on? We didn't encounter this in our pre-production setup.

    Read the article

  • Grub Autostart with timeout

    - by BetaRide
    On Ubuntu 10.4 LTS I want Grub to start the default OS after 5 Seconds. I'd like to see the output of the startup scripts. Currently grub wait forever until I hit return and the output of the startup scripts isn't visible. Can someone tell me how I have to configure /etc/default/grub or any other setups? I tried to play with GRUB_TIMEOUT and GRUB_DEFAULT and did a sudo update-grub afterwards, but nothing changed. Any ideas? # If you change this file, run 'update-grub' afterwards to update # /boot/grub/grub.cfg. GRUB_DEFAULT=0 GRUB_HIDDEN_TIMEOUT=5 #GRUB_HIDDEN_TIMEOUT_QUIET=true GRUB_TIMEOUT=5 GRUB_DISTRIBUTOR=`lsb_release -i -s 2> /dev/null || echo Debian` GRUB_CMDLINE_LINUX_DEFAULT="quiet splash" GRUB_CMDLINE_LINUX="" GRUB_SAVEDEFAULT=true # Uncomment to disable graphical terminal (grub-pc only) #GRUB_TERMINAL=console # The resolution used on graphical terminal # note that you can use only modes which your graphic card supports via VBE # you can see them in real GRUB with the command `vbeinfo' #GRUB_GFXMODE=640x480 # Uncomment if you don't want GRUB to pass "root=UUID=xxx" parameter to Linux # GRUB_DISABLE_LINUX_UUID=true # Uncomment to disable generation of recovery mode menu entries #GRUB_DISABLE_LINUX_RECOVERY="true" # Uncomment to get a beep at grub start #GRUB_INIT_TUNE="480 440 1"

    Read the article

  • Windows 7 delayed file delete

    - by GregoryM
    I'm stuck with a pretty rare problem that happens on Windows 7 OS only. Every time I'm deleting the file with *.exe extension through explorer, the file doesn't get deleted immediately. I'm forced to wait for around one-two minutes before the system will delete the file. The main problem is that I cannot develop in such situation, because every time I build my solution, the old executable gets 'deleted', but is still there. So the new one cannot be created by Visual Studio. This problem breaks the Steam update progress and a few other installers functionality too. Fresh installed Win7 doesn't have this kind of trouble, so I guess this must be some bad registry entries or some services. Browsing the internet for solutions I found only this: http://www.sevenforums.com/software/72091-several-minute-delay-when-deleting-any-exe-file.html. But the solution the author found is not working (change the userName :)). Is there any ideas how to find what causes this to happen? BTW: when I place the file into Trash bin, no delay occurs. When I delete file with Total Commander - no delay too. Tech details: Windows 7 x64 Ultimate. UPD: maybe some shadow copying or system restore services (though I have the system restore turned off) block the files? Can't even guess...

    Read the article

  • Executing a git command using remote powershell results in a NativeCommmandError

    - by user204777
    I am getting an error while executing a remote PowerShell script. From my local machine I am running a PowerShell script that uses Invoke-Command to cd into a directory on a remote Amazon Windows Server instance, and a subsequent Invoke-Command to execute script that lives on that server instance. The script on the server is trying to git clone a repository from GitHub. I can successfully do things in the server script like "ls" or even "git --version". However git clone, git pull, etc. result in the following error: Cloning into 'MyRepo'... + CategoryInfo : NotSpecified: (Cloning into 'MyRepo'...:String) [], RemoteException + FullyQualifiedErrorId : NativeCommandError This is my first time using PowerShell or a Windows Server. Can anyone provide some direction on this problem. The client script: $s = new-pssession -computername $server -credential $user invoke-command -session $s -scriptblock { cd C:\Repos; ls } invoke-command -session $s -scriptblock { param ($repo, $branch) & '.\clone.ps1' -repository $repo -branch $branch} -ArgumentList $repository, $branch exit-pssession The server script: param([string]$repository = "repository", [string]$branch = "branch") git --version start-process -FilePath git -ArgumentList ("clone", "-b $branch https://github.com/MyGithub/$repository.git") -Wait I've changed the server script to use start process and it is no longer throwing the exception. It creates the new repository directory and the .git directory but doesn't write any of the files from the github repository. This smells like a permissions issue. Once again invoking the script manually (remote desktop into the amazon box and execute it from powershell) works like a charm.

    Read the article

  • Mail server DNS failed to resolved by Mac clients

    - by Concordus Applications
    We have two internal DNS servers. One is located on a linux server box and the other is the router's DNS management. We set the linux box as primary DNS via DHCP and the router as secondary. We have a few Mac clients that are accessing our internal mail server (hostnamed "mail" internally). When using IMAP or SMTP against the mail server internally, the mac boxes will sometimes fail to locate the server. If I use NSLOOKUP I can see that "mail" is pointed to the correct IP address and is being resolved via the correct DNS server, but if I ping "mail" it fails. ~ (bash)$ nslookup mail Server: 254.254.254.206 Address: 254.254.254.206#53 Name: mail.example.com Address: 254.254.254.205 Note: I replaced our actual internal IP address with 254.254.254.* If I wait a few minutes (3-5 minutes), somehow it resolves itself and sends successfully. This happens multiple times a day. The /etc/hosts file on the mac boxes is the default config. ## # Host Database # # localhost is used to configure the loopback interface # when the system is booting. Do not change this entry. ## 127.0.0.1 localhost 255.255.255.255 broadcasthost ::1 localhost fe80::1%lo0 localhost Is there something about Mac clients I should know to prevent this failed DNS resolution? Client boxes are: OSX 10.7.4, 8GB RAM, i5 MacBooks Server is: Ubuntu 12.04 Server

    Read the article

  • Some pages begin to load and stop on Chrome

    - by corsiKa
    I'm using Chrome (Version 22.0.1229.94). About 20% of pages simply stop loading after a second or two. Sometimes, I'm able to click an early link after a second or third reload. If I attempt to close the page, it continues to hang for a few (10-30) additional seconds, then closes. However, if I switch to other tabs, it works just fine. If I don't change tabs and wait long enough, it says that there's something on the page that's taking too long to run and offers to let me kill it. Only a select number of sites fail to load, and they do so consistently. None of the stackexchange sites or google fail, but others like realclearpolitics and wowwiki do. I visit those sites every day, and this is the first time it has failed like this. If it were just one site, I would say someone messed something up in their deployment. But it seems incredibly peculiar that suddenly, half a dozen popular sites all mysteriously have the same symptoms. If I attempt to load the pages in Firefox or IE9, they load just fine. Nothing new has been installed, regarding Chrome or otherwise. Antivirus reports no abnormalities. System is regularly patched. Restarting both chrome and the computer have had no effect.

    Read the article

  • Refresh NFS mount

    - by HayekSplosives
    If I check an NFS share on a machine and ls I get the folders. If I got to the NFS host and add a new directory to /etc/exports for the client and do exportfs -a what do I run on the client to refresh the directories? Example (pseudoish): nfsNode01: echo "/share clientNode01 >> /etc/exports"; exportfs -a; clientNode01: cd /share; ls; nfsNode01: echo "/share/folder clientNode01 >> /etc/exports"; exportfs -a; clientNode01: ls; Results as still the same as above. If I reboot the /share/folder shares are there. I know there has to be a way to refresh that info from NFS. I'm sure if I let the connection wait long enough the next time I mounted /dsl would do it. Can I just umount/mount or is there a better way?

    Read the article

  • socket operation on nonsocket or bad file descriptor

    - by Magn3s1um
    I'm writing a pthread server which takes requests from clients and sends them back a bunch of .ppm files. Everything seems to go well, but sometimes when I have just 1 client connected, when trying to read from the file descriptor (for the file), it says Bad file Descriptor. This doesn't make sense, since my int fd isn't -1, and the file most certainly exists. Other times, I get this "Socket operation on nonsocket" error. This is weird because other times, it doesn't give me this error and everything works fine. When trying to connect multiple clients, for some reason, it will only send correctly to one, and then the other client gets the bad file descriptor or "nonsocket" error, even though both threads are processing the same messages and do the same routines. Anyone have an idea why? Here's the code that is giving me that error: while(mqueue.head != mqueue.tail && count < dis_m){ printf("Sending to client %s: %s\n", pointer->id, pointer->message); int fd; fd = open(pointer->message, O_RDONLY); char buf[58368]; int bytesRead; printf("This is fd %d\n", fd); bytesRead=read(fd,buf,58368); send(pointer->socket,buf,bytesRead,0); perror("Error:\n"); fflush(stdout); close(fd); mqueue.mcount--; mqueue.head = mqueue.head->next; free(pointer->message); free(pointer); pointer = mqueue.head; count++; } printf("Sending %s\n", pointer->message); int fd; fd = open(pointer->message, O_RDONLY); printf("This is fd %d\n", fd); printf("I am hhere2\n"); char buf[58368]; int bytesRead; bytesRead=read(fd,buf,58368); send(pointer->socket,buf,bytesRead,0); perror("Error:\n"); close(fd); mqueue.mcount--; if(mqueue.head != mqueue.tail){ mqueue.head = mqueue.head->next; } else{ mqueue.head->next = malloc(sizeof(struct message)); mqueue.head = mqueue.head->next; mqueue.head->next = malloc(sizeof(struct message)); mqueue.tail = mqueue.head->next; mqueue.head->message = NULL; } free(pointer->message); free(pointer); pthread_mutex_unlock(&numm); pthread_mutex_unlock(&circ); pthread_mutex_unlock(&slots); The messages for both threads are the same, being of the form ./path/imageXX.ppm where XX is the number that should go to the client. The file size of each image is 58368 bytes. Sometimes, this code hangs on the read, and stops execution. I don't know this would be either, because the file descriptor comes back as valid. Thanks in advanced. Edit: Here's some sample output: Sending to client a: ./support/images/sw90.ppm This is fd 4 Error: : Socket operation on non-socket Sending to client a: ./support/images/sw91.ppm This is fd 4 Error: : Socket operation on non-socket Sending ./support/images/sw92.ppm This is fd 4 I am hhere2 Error: : Socket operation on non-socket My dispatcher has defeated evil Sample with 2 clients (client b was serviced first) Sending to client b: ./support/images/sw87.ppm This is fd 6 Error: : Success Sending to client b: ./support/images/sw88.ppm This is fd 6 Error: : Success Sending to client b: ./support/images/sw89.ppm This is fd 6 Error: : Success This is fd 6 Error: : Bad file descriptor Sending to client a: ./support/images/sw85.ppm This is fd 6 Error: As you can see, who ever is serviced first in this instance can open the files, but not the 2nd person. Edit2: Full code. Sorry, its pretty long and terribly formatted. #include <netinet/in.h> #include <netinet/in.h> #include <netdb.h> #include <arpa/inet.h> #include <sys/types.h> #include <sys/socket.h> #include <errno.h> #include <stdio.h> #include <unistd.h> #include <pthread.h> #include <stdlib.h> #include <string.h> #include <sys/types.h> #include <sys/stat.h> #include <fcntl.h> #include "ring.h" /* Version 1 Here is what is implemented so far: The threads are created from the arguments specified (number of threads that is) The server will lock and update variables based on how many clients are in the system and such. The socket that is opened when a new client connects, must be passed to the threads. To do this, we need some sort of global array. I did this by specifying an int client and main_pool_busy, and two pointers poolsockets and nonpoolsockets. My thinking on this was that when a new client enters the system, the server thread increments the variable client. When a thread is finished with this client (after it sends it the data), the thread will decrement client and close the socket. HTTP servers act this way sometimes (they terminate the socket as soon as one transmission is sent). *Note down at bottom After the server portion increments the client counter, we must open up a new socket (denoted by new_sd) and get this value to the appropriate thread. To do this, I created global array poolsockets, which will hold all the socket descriptors for our pooled threads. The server portion gets the new socket descriptor, and places the value in the first spot of the array that has a 0. We only place a value in this array IF: 1. The variable main_pool_busy < worknum (If we have more clients in the system than in our pool, it doesn't mean we should always create a new thread. At the end of this, the server signals on the condition variable clientin that a new client has arrived. In our pooled thread, we then must walk this array and check the array until we hit our first non-zero value. This is the socket we will give to that thread. The thread then changes the array to have a zero here. What if our all threads in our pool our busy? If this is the case, then we will know it because our threads in this pool will increment main_pool_busy by one when they are working on a request and decrement it when they are done. If main_pool_busy >= worknum, then we must dynamically create a new thread. Then, we must realloc the size of our nonpoolsockets array by 1 int. We then add the new socket descriptor to our pool. Here's what we need to figure out: NOTE* Each worker should generate 100 messages which specify the worker thread ID, client socket descriptor and a copy of the client message. Additionally, each message should include a message number, starting from 0 and incrementing for each subsequent message sent to the same client. I don't know how to keep track of how many messages were to the same client. Maybe we shouldn't close the socket descriptor, but rather keep an array of structs for each socket that includes how many messages they have been sent. Then, the server adds the struct, the threads remove it, then the threads add it back once they've serviced one request (unless the count is 100). ------------------------------------------------------------- CHANGES Version 1 ---------- NONE: this is the first version. */ #define MAXSLOTS 30 #define dis_m 15 //problems with dis_m ==1 //Function prototypes void inc_clients(); void init_mutex_stuff(pthread_t*, pthread_t*); void *threadpool(void *); void server(int); void add_to_socket_pool(int); void inc_busy(); void dec_busy(); void *dispatcher(); void create_message(long, int, int, char *, char *); void init_ring(); void add_to_ring(char *, char *, int, int, int); int socket_from_string(char *); void add_to_head(char *); void add_to_tail(char *); struct message * reorder(struct message *, struct message *, int); int get_threadid(char *); void delete_socket_messages(int); struct message * merge(struct message *, struct message *, int); int get_request(char *, char *, char*); ///////////////////// //Global mutexes and condition variables pthread_mutex_t startservice; pthread_mutex_t numclients; pthread_mutex_t pool_sockets; pthread_mutex_t nonpool_sockets; pthread_mutex_t m_pool_busy; pthread_mutex_t slots; pthread_mutex_t numm; pthread_mutex_t circ; pthread_cond_t clientin; pthread_cond_t m; /////////////////////////////////////// //Global variables int clients; int main_pool_busy; int * poolsockets, nonpoolsockets; int worknum; struct ring mqueue; /////////////////////////////////////// int main(int argc, char ** argv){ //error handling if not enough arguments to program if(argc != 3){ printf("Not enough arguments to server: ./server portnum NumThreadsinPool\n"); _exit(-1); } //Convert arguments from strings to integer values int port = atoi(argv[1]); worknum = atoi(argv[2]); //Start server portion server(port); } /////////////////////////////////////////////////////////////////////////////////////////////// //The listen server thread///////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////////////////////////////// void server(int port){ int sd, new_sd; struct sockaddr_in name, cli_name; int sock_opt_val = 1; int cli_len; pthread_t threads[worknum]; //create our pthread id array pthread_t dis[1]; //create our dispatcher array (necessary to create thread) init_mutex_stuff(threads, dis); //initialize mutexes and stuff //Server setup /////////////////////////////////////////////////////// if ((sd = socket (AF_INET, SOCK_STREAM, 0)) < 0) { perror("(servConn): socket() error"); _exit (-1); } if (setsockopt (sd, SOL_SOCKET, SO_REUSEADDR, (char *) &sock_opt_val, sizeof(sock_opt_val)) < 0) { perror ("(servConn): Failed to set SO_REUSEADDR on INET socket"); _exit (-1); } name.sin_family = AF_INET; name.sin_port = htons (port); name.sin_addr.s_addr = htonl(INADDR_ANY); if (bind (sd, (struct sockaddr *)&name, sizeof(name)) < 0) { perror ("(servConn): bind() error"); _exit (-1); } listen (sd, 5); //End of server Setup ////////////////////////////////////////////////// for (;;) { cli_len = sizeof (cli_name); new_sd = accept (sd, (struct sockaddr *) &cli_name, &cli_len); printf ("Assigning new socket descriptor: %d\n", new_sd); inc_clients(); //New client has come in, increment clients add_to_socket_pool(new_sd); //Add client to the pool of sockets if (new_sd < 0) { perror ("(servConn): accept() error"); _exit (-1); } } pthread_exit(NULL); //Quit } //Adds the new socket to the array designated for pthreads in the pool void add_to_socket_pool(int socket){ pthread_mutex_lock(&m_pool_busy); //Lock so that we can check main_pool_busy int i; //If not all our main pool is busy, then allocate to one of them if(main_pool_busy < worknum){ pthread_mutex_unlock(&m_pool_busy); //unlock busy, we no longer need to hold it pthread_mutex_lock(&pool_sockets); //Lock the socket pool array so that we can edit it without worry for(i = 0; i < worknum; i++){ //Find a poolsocket that is -1; then we should put the real socket there. This value will be changed back to -1 when the thread grabs the sockfd if(poolsockets[i] == -1){ poolsockets[i] = socket; pthread_mutex_unlock(&pool_sockets); //unlock our pool array, we don't need it anymore inc_busy(); //Incrememnt busy (locks the mutex itself) pthread_cond_signal(&clientin); //Signal first thread waiting on a client that a client needs to be serviced break; } } } else{ //Dynamic thread creation goes here pthread_mutex_unlock(&m_pool_busy); } } //Increments the client number. If client number goes over worknum, we must dynamically create new pthreads void inc_clients(){ pthread_mutex_lock(&numclients); clients++; pthread_mutex_unlock(&numclients); } //Increments busy void inc_busy(){ pthread_mutex_lock(&m_pool_busy); main_pool_busy++; pthread_mutex_unlock(&m_pool_busy); } //Initialize all of our mutexes at the beginning and create our pthreads void init_mutex_stuff(pthread_t * threads, pthread_t * dis){ pthread_mutex_init(&startservice, NULL); pthread_mutex_init(&numclients, NULL); pthread_mutex_init(&pool_sockets, NULL); pthread_mutex_init(&nonpool_sockets, NULL); pthread_mutex_init(&m_pool_busy, NULL); pthread_mutex_init(&circ, NULL); pthread_cond_init (&clientin, NULL); main_pool_busy = 0; poolsockets = malloc(sizeof(int)*worknum); int threadreturn; //error checking variables long i = 0; //Loop and create pthreads for(i; i < worknum; i++){ threadreturn = pthread_create(&threads[i], NULL, threadpool, (void *) i); poolsockets[i] = -1; if(threadreturn){ perror("Thread pool created unsuccessfully"); _exit(-1); } } pthread_create(&dis[0], NULL, dispatcher, NULL); } ////////////////////////////////////////////////////////////////////////////////////////// /////////Main pool routines ///////////////////////////////////////////////////////////////////////////////////////// void dec_busy(){ pthread_mutex_lock(&m_pool_busy); main_pool_busy--; pthread_mutex_unlock(&m_pool_busy); } void dec_clients(){ pthread_mutex_lock(&numclients); clients--; pthread_mutex_unlock(&numclients); } //This is what our threadpool pthreads will be running. void *threadpool(void * threadid){ long id = (long) threadid; //Id of this thread int i; int socket; int counter = 0; //Try and gain access to the next client that comes in and wait until server signals that a client as arrived while(1){ pthread_mutex_lock(&startservice); //lock start service (required for cond wait) pthread_cond_wait(&clientin, &startservice); //wait for signal from server that client exists pthread_mutex_unlock(&startservice); //unlock mutex. pthread_mutex_lock(&pool_sockets); //Lock the pool socket so we can get the socket fd unhindered/interrupted for(i = 0; i < worknum; i++){ if(poolsockets[i] != -1){ socket = poolsockets[i]; poolsockets[i] = -1; pthread_mutex_unlock(&pool_sockets); } } printf("Thread #%d is past getting the socket\n", id); int incoming = 1; while(counter < 100 && incoming != 0){ char buffer[512]; bzero(buffer,512); int startcounter = 0; incoming = read(socket, buffer, 512); if(buffer[0] != 0){ //client ID:priority:request:arguments char id[100]; long prior; char request[100]; char arg1[100]; char message[100]; char arg2[100]; char * point; point = strtok(buffer, ":"); strcpy(id, point); point = strtok(NULL, ":"); prior = atoi(point); point = strtok(NULL, ":"); strcpy(request, point); point = strtok(NULL, ":"); strcpy(arg1, point); point = strtok(NULL, ":"); if(point != NULL){ strcpy(arg2, point); } int fd; if(strcmp(request, "start_movie") == 0){ int count = 1; while(count <= 100){ char temp[10]; snprintf(temp, 50, "%d\0", count); strcpy(message, "./support/images/"); strcat(message, arg1); strcat(message, temp); strcat(message, ".ppm"); printf("This is message %s to %s\n", message, id); count++; add_to_ring(message, id, prior, counter, socket); //Adds our created message to the ring counter++; } printf("I'm out of the loop\n"); } else if(strcmp(request, "seek_movie") == 0){ int count = atoi(arg2); while(count <= 100){ char temp[10]; snprintf(temp, 10, "%d\0", count); strcpy(message, "./support/images/"); strcat(message, arg1); strcat(message, temp); strcat(message, ".ppm"); printf("This is message %s\n", message); count++; } } //create_message(id, socket, counter, buffer, message); //Creates our message from the input from the client. Stores it in buffer } else{ delete_socket_messages(socket); break; } } counter = 0; close(socket);//Zero out counter again } dec_clients(); //client serviced, decrement clients dec_busy(); //thread finished, decrement busy } //Creates a message void create_message(long threadid, int socket, int counter, char * buffer, char * message){ snprintf(message, strlen(buffer)+15, "%d:%d:%d:%s", threadid, socket, counter, buffer); } //Gets the socket from the message string (maybe I should just pass in the socket to another method) int socket_from_string(char * message){ char * substr1 = strstr(message, ":"); char * substr2 = substr1; substr2++; int occurance = strcspn(substr2, ":"); char sock[10]; strncpy(sock, substr2, occurance); return atoi(sock); } //Adds message to our ring buffer's head void add_to_head(char * message){ printf("Adding to head of ring\n"); mqueue.head->message = malloc(strlen(message)+1); //Allocate space for message strcpy(mqueue.head->message, message); //copy bytes into allocated space } //Adds our message to our ring buffer's tail void add_to_tail(char * message){ printf("Adding to tail of ring\n"); mqueue.tail->message = malloc(strlen(message)+1); //allocate space for message strcpy(mqueue.tail->message, message); //copy bytes into allocated space mqueue.tail->next = malloc(sizeof(struct message)); //allocate space for the next message struct } //Adds a message to our ring void add_to_ring(char * message, char * id, int prior, int mnum, int socket){ //printf("This is message %s:" , message); pthread_mutex_lock(&circ); //Lock the ring buffer pthread_mutex_lock(&numm); //Lock the message count (will need this to make sure we can't fill the buffer over the max slots) if(mqueue.head->message == NULL){ add_to_head(message); //Adds it to head mqueue.head->socket = socket; //Set message socket mqueue.head->priority = prior; //Set its priority (thread id) mqueue.head->mnum = mnum; //Set its message number (used for sorting) mqueue.head->id = malloc(sizeof(id)); strcpy(mqueue.head->id, id); } else if(mqueue.tail->message == NULL){ //This is the problem for dis_m 1 I'm pretty sure add_to_tail(message); mqueue.tail->socket = socket; mqueue.tail->priority = prior; mqueue.tail->mnum = mnum; mqueue.tail->id = malloc(sizeof(id)); strcpy(mqueue.tail->id, id); } else{ mqueue.tail->next = malloc(sizeof(struct message)); mqueue.tail = mqueue.tail->next; add_to_tail(message); mqueue.tail->socket = socket; mqueue.tail->priority = prior; mqueue.tail->mnum = mnum; mqueue.tail->id = malloc(sizeof(id)); strcpy(mqueue.tail->id, id); } mqueue.mcount++; pthread_mutex_unlock(&circ); if(mqueue.mcount >= dis_m){ pthread_mutex_unlock(&numm); pthread_cond_signal(&m); } else{ pthread_mutex_unlock(&numm); } printf("out of add to ring\n"); fflush(stdout); } ////////////////////////////////// //Dispatcher routines ///////////////////////////////// void *dispatcher(){ init_ring(); while(1){ pthread_mutex_lock(&slots); pthread_cond_wait(&m, &slots); pthread_mutex_lock(&numm); pthread_mutex_lock(&circ); printf("Dispatcher to the rescue!\n"); mqueue.head = reorder(mqueue.head, mqueue.tail, mqueue.mcount); //printf("This is the head %s\n", mqueue.head->message); //printf("This is the tail %s\n", mqueue.head->message); fflush(stdout); struct message * pointer = mqueue.head; int count = 0; while(mqueue.head != mqueue.tail && count < dis_m){ printf("Sending to client %s: %s\n", pointer->id, pointer->message); int fd; fd = open(pointer->message, O_RDONLY); char buf[58368]; int bytesRead; printf("This is fd %d\n", fd); bytesRead=read(fd,buf,58368); send(pointer->socket,buf,bytesRead,0); perror("Error:\n"); fflush(stdout); close(fd); mqueue.mcount--; mqueue.head = mqueue.head->next; free(pointer->message); free(pointer); pointer = mqueue.head; count++; } printf("Sending %s\n", pointer->message); int fd; fd = open(pointer->message, O_RDONLY); printf("This is fd %d\n", fd); printf("I am hhere2\n"); char buf[58368]; int bytesRead; bytesRead=read(fd,buf,58368); send(pointer->socket,buf,bytesRead,0); perror("Error:\n"); close(fd); mqueue.mcount--; if(mqueue.head != mqueue.tail){ mqueue.head = mqueue.head->next; } else{ mqueue.head->next = malloc(sizeof(struct message)); mqueue.head = mqueue.head->next; mqueue.head->next = malloc(sizeof(struct message)); mqueue.tail = mqueue.head->next; mqueue.head->message = NULL; } free(pointer->message); free(pointer); pthread_mutex_unlock(&numm); pthread_mutex_unlock(&circ); pthread_mutex_unlock(&slots); printf("My dispatcher has defeated evil\n"); } } void init_ring(){ mqueue.head = malloc(sizeof(struct message)); mqueue.head->next = malloc(sizeof(struct message)); mqueue.tail = mqueue.head->next; mqueue.mcount = 0; } struct message * reorder(struct message * begin, struct message * end, int num){ //printf("I am reordering for size %d\n", num); fflush(stdout); int i; if(num == 1){ //printf("Begin: %s\n", begin->message); begin->next = NULL; return begin; } else{ struct message * left = begin; struct message * right; int middle = num/2; for(i = 1; i < middle; i++){ left = left->next; } right = left -> next; left -> next = NULL; //printf("Begin: %s\nLeft: %s\nright: %s\nend:%s\n", begin->message, left->message, right->message, end->message); left = reorder(begin, left, middle); if(num%2 != 0){ right = reorder(right, end, middle+1); } else{ right = reorder(right, end, middle); } return merge(left, right, num); } } struct message * merge(struct message * left, struct message * right, int num){ //printf("I am merginging! left: %s %d, right: %s %dnum: %d\n", left->message,left->priority, right->message, right->priority, num); struct message * start, * point; int lenL= 0; int lenR = 0; int flagL = 0; int flagR = 0; int count = 0; int middle1 = num/2; int middle2; if(num%2 != 0){ middle2 = middle1+1; } else{ middle2 = middle1; } while(lenL < middle1 && lenR < middle2){ count++; //printf("In here for count %d\n", count); if(lenL == 0 && lenR == 0){ if(left->priority < right->priority){ start = left; //Set the start point point = left; //set our enum; left = left->next; //move the left pointer point->next = NULL; //Set the next node to NULL lenL++; } else if(left->priority > right->priority){ start = right; point = right; right = right->next; point->next = NULL; lenR++; } else{ if(left->mnum < right->mnum){ ////printf("This is where we are\n"); start = left; //Set the start point point = left; //set our enum; left = left->next; //move the left pointer point->next = NULL; //Set the next node to NULL lenL++; } else{ start = right; point = right; right = right->next; point->next = NULL; lenR++; } } } else{ if(left->priority < right->priority){ point->next = left; left = left->next; //move the left pointer point = point->next; point->next = NULL; //Set the next node to NULL lenL++; } else if(left->priority > right->priority){ point->next = right; right = right->next; point = point->next; point->next = NULL; lenR++; } else{ if(left->mnum < right->mnum){ point->next = left; //set our enum; left = left->next; point = point->next;//move the left pointer point->next = NULL; //Set the next node to NULL lenL++; } else{ point->next = right; right = right->next; point = point->next; point->next = NULL; lenR++; } } } if(lenL == middle1){ flagL = 1; break; } if(lenR == middle2){ flagR = 1; break; } } if(flagL == 1){ point->next = right; point = point->next; for(lenR; lenR< middle2-1; lenR++){ point = point->next; } point->next = NULL; mqueue.tail = point; } else{ point->next = left; point = point->next; for(lenL; lenL< middle1-1; lenL++){ point = point->next; } point->next = NULL; mqueue.tail = point; } //printf("This is the start %s\n", start->message); //printf("This is mqueue.tail %s\n", mqueue.tail->message); return start; } void delete_socket_messages(int a){ }

    Read the article

  • Red Gate Coder interviews: Alex Davies

    - by Michael Williamson
    Alex Davies has been a software engineer at Red Gate since graduating from university, and is currently busy working on .NET Demon. We talked about tackling parallel programming with his actors framework, a scientific approach to debugging, and how JavaScript is going to affect the programming languages we use in years to come. So, if we start at the start, how did you get started in programming? When I was seven or eight, I was given a BBC Micro for Christmas. I had asked for a Game Boy, but my dad thought it would be better to give me a proper computer. For a year or so, I only played games on it, but then I found the user guide for writing programs in it. I gradually started doing more stuff on it and found it fun. I liked creating. As I went into senior school I continued to write stuff on there, trying to write games that weren’t very good. I got a real computer when I was fourteen and found ways to write BASIC on it. Visual Basic to start with, and then something more interesting than that. How did you learn to program? Was there someone helping you out? Absolutely not! I learnt out of a book, or by experimenting. I remember the first time I found a loop, I was like “Oh my God! I don’t have to write out the same line over and over and over again any more. It’s amazing!” When did you think this might be something that you actually wanted to do as a career? For a long time, I thought it wasn’t something that you would do as a career, because it was too much fun to be a career. I thought I’d do chemistry at university and some kind of career based on chemical engineering. And then I went to a careers fair at school when I was seventeen or eighteen, and it just didn’t interest me whatsoever. I thought “I could be a programmer, and there’s loads of money there, and I’m good at it, and it’s fun”, but also that I shouldn’t spoil my hobby. Now I don’t really program in my spare time any more, which is a bit of a shame, but I program all the rest of the time, so I can live with it. Do you think you learnt much about programming at university? Yes, definitely! I went into university knowing how to make computers do anything I wanted them to do. However, I didn’t have the language to talk about algorithms, so the algorithms course in my first year was massively important. Learning other language paradigms like functional programming was really good for breadth of understanding. Functional programming influences normal programming through design rather than actually using it all the time. I draw inspiration from it to write imperative programs which I think is actually becoming really fashionable now, but I’ve been doing it for ages. I did it first! There were also some courses on really odd programming languages, a bit of Prolog, a little bit of C. Having a little bit of each of those is something that I would have never done on my own, so it was important. And then there are knowledge-based courses which are about not programming itself but things that have been programmed like TCP. Those are really important for examples for how to approach things. Did you do any internships while you were at university? Yeah, I spent both of my summers at the same company. I thought I could code well before I went there. Looking back at the crap that I produced, it was only surpassed in its crappiness by all of the other code already in that company. I’m so much better at writing nice code now than I used to be back then. Was there just not a culture of looking after your code? There was, they just didn’t hire people for their abilities in that area. They hired people for raw IQ. The first indicator of it going wrong was that they didn’t have any computer scientists, which is a bit odd in a programming company. But even beyond that they didn’t have people who learnt architecture from anyone else. Most of them had started straight out of university, so never really had experience or mentors to learn from. There wasn’t the experience to draw from to teach each other. In the second half of my second internship, I was being given tasks like looking at new technologies and teaching people stuff. Interns shouldn’t be teaching people how to do their jobs! All interns are going to have little nuggets of things that you don’t know about, but they shouldn’t consistently be the ones who know the most. It’s not a good environment to learn. I was going to ask how you found working with people who were more experienced than you… When I reached Red Gate, I found some people who were more experienced programmers than me, and that was difficult. I’ve been coding since I was tiny. At university there were people who were cleverer than me, but there weren’t very many who were more experienced programmers than me. During my internship, I didn’t find anyone who I classed as being a noticeably more experienced programmer than me. So, it was a shock to the system to have valid criticisms rather than just formatting criticisms. However, Red Gate’s not so big on the actual code review, at least it wasn’t when I started. We did an entire product release and then somebody looked over all of the UI of that product which I’d written and say what they didn’t like. By that point, it was way too late and I’d disagree with them. Do you think the lack of code reviews was a bad thing? I think if there’s going to be any oversight of new people, then it should be continuous rather than chunky. For me I don’t mind too much, I could go out and get oversight if I wanted it, and in those situations I felt comfortable without it. If I was managing the new person, then maybe I’d be keener on oversight and then the right way to do it is continuously and in very, very small chunks. Have you had any significant projects you’ve worked on outside of a job? When I was a teenager I wrote all sorts of stuff. I used to write games, I derived how to do isomorphic projections myself once. I didn’t know what the word was so I couldn’t Google for it, so I worked it out myself. It was horrifically complicated. But it sort of tailed off when I started at university, and is now basically zero. If I do side-projects now, they tend to be work-related side projects like my actors framework, NAct, which I started in a down tools week. Could you explain a little more about NAct? It is a little C# framework for writing parallel code more easily. Parallel programming is difficult when you need to write to shared data. Sometimes parallel programming is easy because you don’t need to write to shared data. When you do need to access shared data, you could just have your threads pile in and do their work, but then you would screw up the data because the threads would trample on each other’s toes. You could lock, but locks are really dangerous if you’re using more than one of them. You get interactions like deadlocks, and that’s just nasty. Actors instead allows you to say this piece of data belongs to this thread of execution, and nobody else can read it. If you want to read it, then ask that thread of execution for a piece of it by sending a message, and it will send the data back by a message. And that avoids deadlocks as long as you follow some obvious rules about not making your actors sit around waiting for other actors to do something. There are lots of ways to write actors, NAct allows you to do it as if it was method calls on other objects, which means you get all the strong type-safety that C# programmers like. Do you think that this is suitable for the majority of parallel programming, or do you think it’s only suitable for specific cases? It’s suitable for most difficult parallel programming. If you’ve just got a hundred web requests which are all independent of each other, then I wouldn’t bother because it’s easier to just spin them up in separate threads and they can proceed independently of each other. But where you’ve got difficult parallel programming, where you’ve got multiple threads accessing multiple bits of data in multiple ways at different times, then actors is at least as good as all other ways, and is, I reckon, easier to think about. When you’re using actors, you presumably still have to write your code in a different way from you would otherwise using single-threaded code. You can’t use actors with any methods that have return types, because you’re not allowed to call into another actor and wait for it. If you want to get a piece of data out of another actor, then you’ve got to use tasks so that you can use “async” and “await” to await asynchronously for it. But other than that, you can still stick things in classes so it’s not too different really. Rather than having thousands of objects with mutable state, you can use component-orientated design, where there are only a few mutable classes which each have a small number of instances. Then there can be thousands of immutable objects. If you tend to do that anyway, then actors isn’t much of a jump. If I’ve already built my system without any parallelism, how hard is it to add actors to exploit all eight cores on my desktop? Usually pretty easy. If you can identify even one boundary where things look like messages and you have components where some objects live on one side and these other objects live on the other side, then you can have a granddaddy object on one side be an actor and it will parallelise as it goes across that boundary. Not too difficult. If we do get 1000-core desktop PCs, do you think actors will scale up? It’s hard. There are always in the order of twenty to fifty actors in my whole program because I tend to write each component as actors, and I tend to have one instance of each component. So this won’t scale to a thousand cores. What you can do is write data structures out of actors. I use dictionaries all over the place, and if you need a dictionary that is going to be accessed concurrently, then you could build one of those out of actors in no time. You can use queuing to marshal requests between different slices of the dictionary which are living on different threads. So it’s like a distributed hash table but all of the chunks of it are on the same machine. That means that each of these thousand processors has cached one small piece of the dictionary. I reckon it wouldn’t be too big a leap to start doing proper parallelism. Do you think it helps if actors get baked into the language, similarly to Erlang? Erlang is excellent in that it has thread-local garbage collection. C# doesn’t, so there’s a limit to how well C# actors can possibly scale because there’s a single garbage collected heap shared between all of them. When you do a global garbage collection, you’ve got to stop all of the actors, which is seriously expensive, whereas in Erlang garbage collections happen per-actor, so they’re insanely cheap. However, Erlang deviated from all the sensible language design that people have used recently and has just come up with crazy stuff. You can definitely retrofit thread-local garbage collection to .NET, and then it’s quite well-suited to support actors, even if it’s not baked into the language. Speaking of language design, do you have a favourite programming language? I’ll choose a language which I’ve never written before. I like the idea of Scala. It sounds like C#, only with some of the niggles gone. I enjoy writing static types. It means you don’t have to writing tests so much. When you say it doesn’t have some of the niggles? C# doesn’t allow the use of a property as a method group. It doesn’t have Scala case classes, or sum types, where you can do a switch statement and the compiler checks that you’ve checked all the cases, which is really useful in functional-style programming. Pattern-matching, in other words. That’s actually the major niggle. C# is pretty good, and I’m quite happy with C#. And what about going even further with the type system to remove the need for tests to something like Haskell? Or is that a step too far? I’m quite a pragmatist, I don’t think I could deal with trying to write big systems in languages with too few other users, especially when learning how to structure things. I just don’t know anyone who can teach me, and the Internet won’t teach me. That’s the main reason I wouldn’t use it. If I turned up at a company that writes big systems in Haskell, I would have no objection to that, but I wouldn’t instigate it. What about things in C#? For instance, there’s contracts in C#, so you can try to statically verify a bit more about your code. Do you think that’s useful, or just not worthwhile? I’ve not really tried it. My hunch is that it needs to be built into the language and be quite mathematical for it to work in real life, and that doesn’t seem to have ended up true for C# contracts. I don’t think anyone who’s tried them thinks they’re any good. I might be wrong. On a slightly different note, how do you like to debug code? I think I’m quite an odd debugger. I use guesswork extremely rarely, especially if something seems quite difficult to debug. I’ve been bitten spending hours and hours on guesswork and not being scientific about debugging in the past, so now I’m scientific to a fault. What I want is to see the bug happening in the debugger, to step through the bug happening. To watch the program going from a valid state to an invalid state. When there’s a bug and I can’t work out why it’s happening, I try to find some piece of evidence which places the bug in one section of the code. From that experiment, I binary chop on the possible causes of the bug. I suppose that means binary chopping on places in the code, or binary chopping on a stage through a processing cycle. Basically, I’m very stupid about how I debug. I won’t make any guesses, I won’t use any intuition, I will only identify the experiment that’s going to binary chop most effectively and repeat rather than trying to guess anything. I suppose it’s quite top-down. Is most of the time then spent in the debugger? Absolutely, if at all possible I will never debug using print statements or logs. I don’t really hold much stock in outputting logs. If there’s any bug which can be reproduced locally, I’d rather do it in the debugger than outputting logs. And with SmartAssembly error reporting, there’s not a lot that can’t be either observed in an error report and just fixed, or reproduced locally. And in those other situations, maybe I’ll use logs. But I hate using logs. You stare at the log, trying to guess what’s going on, and that’s exactly what I don’t like doing. You have to just look at it and see does this look right or wrong. We’ve covered how you get to grip with bugs. How do you get to grips with an entire codebase? I watch it in the debugger. I find little bugs and then try to fix them, and mostly do it by watching them in the debugger and gradually getting an understanding of how the code works using my process of binary chopping. I have to do a lot of reading and watching code to choose where my slicing-in-half experiment is going to be. The last time I did it was SmartAssembly. The old code was a complete mess, but at least it did things top to bottom. There wasn’t too much of some of the big abstractions where flow of control goes all over the place, into a base class and back again. Code’s really hard to understand when that happens. So I like to choose a little bug and try to fix it, and choose a bigger bug and try to fix it. Definitely learn by doing. I want to always have an aim so that I get a little achievement after every few hours of debugging. Once I’ve learnt the codebase I might be able to fix all the bugs in an hour, but I’d rather be using them as an aim while I’m learning the codebase. If I was a maintainer of a codebase, what should I do to make it as easy as possible for you to understand? Keep distinct concepts in different places. And name your stuff so that it’s obvious which concepts live there. You shouldn’t have some variable that gets set miles up the top of somewhere, and then is read miles down to choose some later behaviour. I’m talking from a very much SmartAssembly point of view because the old SmartAssembly codebase had tons and tons of these things, where it would read some property of the code and then deal with it later. Just thousands of variables in scope. Loads of things to think about. If you can keep concepts separate, then it aids me in my process of fixing bugs one at a time, because each bug is going to more or less be understandable in the one place where it is. And what about tests? Do you think they help at all? I’ve never had the opportunity to learn a codebase which has had tests, I don’t know what it’s like! What about when you’re actually developing? How useful do you find tests in finding bugs or regressions? Finding regressions, absolutely. Running bits of code that would be quite hard to run otherwise, definitely. It doesn’t happen very often that a test finds a bug in the first place. I don’t really buy nebulous promises like tests being a good way to think about the spec of the code. My thinking goes something like “This code works at the moment, great, ship it! Ah, there’s a way that this code doesn’t work. Okay, write a test, demonstrate that it doesn’t work, fix it, use the test to demonstrate that it’s now fixed, and keep the test for future regressions.” The most valuable tests are for bugs that have actually happened at some point, because bugs that have actually happened at some point, despite the fact that you think you’ve fixed them, are way more likely to appear again than new bugs are. Does that mean that when you write your code the first time, there are no tests? Often. The chance of there being a bug in a new feature is relatively unaffected by whether I’ve written a test for that new feature because I’m not good enough at writing tests to think of bugs that I would have written into the code. So not writing regression tests for all of your code hasn’t affected you too badly? There are different kinds of features. Some of them just always work, and are just not flaky, they just continue working whatever you throw at them. Maybe because the type-checker is particularly effective around them. Writing tests for those features which just tend to always work is a waste of time. And because it’s a waste of time I’ll tend to wait until a feature has demonstrated its flakiness by having bugs in it before I start trying to test it. You can get a feel for whether it’s going to be flaky code as you’re writing it. I try to write it to make it not flaky, but there are some things that are just inherently flaky. And very occasionally, I’ll think “this is going to be flaky” as I’m writing, and then maybe do a test, but not most of the time. How do you think your programming style has changed over time? I’ve got clearer about what the right way of doing things is. I used to flip-flop a lot between different ideas. Five years ago I came up with some really good ideas and some really terrible ideas. All of them seemed great when I thought of them, but they were quite diverse ideas, whereas now I have a smaller set of reliable ideas that are actually good for structuring code. So my code is probably more similar to itself than it used to be back in the day, when I was trying stuff out. I’ve got more disciplined about encapsulation, I think. There are operational things like I use actors more now than I used to, and that forces me to use immutability more than I used to. The first code that I wrote in Red Gate was the memory profiler UI, and that was an actor, I just didn’t know the name of it at the time. I don’t really use object-orientation. By object-orientation, I mean having n objects of the same type which are mutable. I want a constant number of objects that are mutable, and they should be different types. I stick stuff in dictionaries and then have one thing that owns the dictionary and puts stuff in and out of it. That’s definitely a pattern that I’ve seen recently. I think maybe I’m doing functional programming. Possibly. It’s plausible. If you had to summarise the essence of programming in a pithy sentence, how would you do it? Programming is the form of art that, without losing any of the beauty of architecture or fine art, allows you to produce things that people love and you make money from. So you think it’s an art rather than a science? It’s a little bit of engineering, a smidgeon of maths, but it’s not science. Like architecture, programming is on that boundary between art and engineering. If you want to do it really nicely, it’s mostly art. You can get away with doing architecture and programming entirely by having a good engineering mind, but you’re not going to produce anything nice. You’re not going to have joy doing it if you’re an engineering mind. Architects who are just engineering minds are not going to enjoy their job. I suppose engineering is the foundation on which you build the art. Exactly. How do you think programming is going to change over the next ten years? There will be an unfortunate shift towards dynamically-typed languages, because of JavaScript. JavaScript has an unfair advantage. JavaScript’s unfair advantage will cause more people to be exposed to dynamically-typed languages, which means other dynamically-typed languages crop up and the best features go into dynamically-typed languages. Then people conflate the good features with the fact that it’s dynamically-typed, and more investment goes into dynamically-typed languages. They end up better, so people use them. What about the idea of compiling other languages, possibly statically-typed, to JavaScript? It’s a reasonable idea. I would like to do it, but I don’t think enough people in the world are going to do it to make it pick up. The hordes of beginners are the lifeblood of a language community. They are what makes there be good tools and what makes there be vibrant community websites. And any particular thing which is the same as JavaScript only with extra stuff added to it, although it might be technically great, is not going to have the hordes of beginners. JavaScript is always to be quickest and easiest way for a beginner to start programming in the browser. And dynamically-typed languages are great for beginners. Compilers are pretty scary and beginners don’t write big code. And having your errors come up in the same place, whether they’re statically checkable errors or not, is quite nice for a beginner. If someone asked me to teach them some programming, I’d teach them JavaScript. If dynamically-typed languages are great for beginners, when do you think the benefits of static typing start to kick in? The value of having a statically typed program is in the tools that rely on the static types to produce a smooth IDE experience rather than actually telling me my compile errors. And only once you’re experienced enough a programmer that having a really smooth IDE experience makes a blind bit of difference, does static typing make a blind bit of difference. So it’s not really about size of codebase. If I go and write up a tiny program, I’m still going to get value out of writing it in C# using ReSharper because I’m experienced with C# and ReSharper enough to be able to write code five times faster if I have that help. Any other visions of the future? Nobody’s going to use actors. Because everyone’s going to be running on single-core VMs connected over network-ready protocols like JSON over HTTP. So, parallelism within one operating system is going to die. But until then, you should use actors. More Red Gater Coder interviews

    Read the article

  • C# .Net 3.5 Asynchronous Socket Server Performance Problem

    - by iBrAaAa
    I'm developing an Asynchronous Game Server using .Net Socket Asynchronous Model( BeginAccept/EndAccept...etc.) The problem I'm facing is described like that: When I have only one client connected, the server response time is very fast but once a second client connects, the server response time increases too much. I've measured the time from a client sends a message to the server until it gets the reply in both cases. I found that the average time in case of one client is about 17ms and in case of 2 clients about 280ms!!! What I really see is that: When 2 clients are connected and only one of them is moving(i.e. requesting service from the server) it is equivalently equal to the case when only one client is connected(i.e. fast response). However, when the 2 clients move at the same time(i.e. requests service from the server at the same time) their motion becomes very slow (as if the server replies each one of them in order i.e. not simultaneously). Basically, what I am doing is that: When a client requests a permission for motion from the server and the server grants him the request, the server then broadcasts the new position of the client to all the players. So if two clients are moving in the same time, the server is eventually trying to broadcast to both clients the new position of each of them at the same time. EX: Client1 asks to go to position (2,2) Client2 asks to go to position (5,5) Server sends to each of Client1 & Client2 the same two messages: message1: "Client1 at (2,2)" message2: "Client2 at (5,5)" I believe that the problem comes from the fact that Socket class is thread safe according MSDN documentation http://msdn.microsoft.com/en-us/library/system.net.sockets.socket.aspx. (NOT SURE THAT IT IS THE PROBLEM) Below is the code for the server: /// /// This class is responsible for handling packet receiving and sending /// public class NetworkManager { /// /// An integer to hold the server port number to be used for the connections. Its default value is 5000. /// private readonly int port = 5000; /// /// hashtable contain all the clients connected to the server. /// key: player Id /// value: socket /// private readonly Hashtable connectedClients = new Hashtable(); /// /// An event to hold the thread to wait for a new client /// private readonly ManualResetEvent resetEvent = new ManualResetEvent(false); /// /// keeps track of the number of the connected clients /// private int clientCount; /// /// The socket of the server at which the clients connect /// private readonly Socket mainSocket = new Socket(AddressFamily.InterNetwork, SocketType.Stream, ProtocolType.Tcp); /// /// The socket exception that informs that a client is disconnected /// private const int ClientDisconnectedErrorCode = 10054; /// /// The only instance of this class. /// private static readonly NetworkManager networkManagerInstance = new NetworkManager(); /// /// A delegate for the new client connected event. /// /// the sender object /// the event args public delegate void NewClientConnected(Object sender, SystemEventArgs e); /// /// A delegate for the position update message reception. /// /// the sender object /// the event args public delegate void PositionUpdateMessageRecieved(Object sender, PositionUpdateEventArgs e); /// /// The event which fires when a client sends a position message /// public PositionUpdateMessageRecieved PositionUpdateMessageEvent { get; set; } /// /// keeps track of the number of the connected clients /// public int ClientCount { get { return clientCount; } } /// /// A getter for this class instance. /// /// only instance. public static NetworkManager NetworkManagerInstance { get { return networkManagerInstance; } } private NetworkManager() {} /// Starts the game server and holds this thread alive /// public void StartServer() { //Bind the mainSocket to the server IP address and port mainSocket.Bind(new IPEndPoint(IPAddress.Any, port)); //The server starts to listen on the binded socket with max connection queue //1024 mainSocket.Listen(1024); //Start accepting clients asynchronously mainSocket.BeginAccept(OnClientConnected, null); //Wait until there is a client wants to connect resetEvent.WaitOne(); } /// /// Receives connections of new clients and fire the NewClientConnected event /// private void OnClientConnected(IAsyncResult asyncResult) { Interlocked.Increment(ref clientCount); ClientInfo newClient = new ClientInfo { WorkerSocket = mainSocket.EndAccept(asyncResult), PlayerId = clientCount }; //Add the new client to the hashtable and increment the number of clients connectedClients.Add(newClient.PlayerId, newClient); //fire the new client event informing that a new client is connected to the server if (NewClientEvent != null) { NewClientEvent(this, System.EventArgs.Empty); } newClient.WorkerSocket.BeginReceive(newClient.Buffer, 0, BasePacket.GetMaxPacketSize(), SocketFlags.None, new AsyncCallback(WaitForData), newClient); //Start accepting clients asynchronously again mainSocket.BeginAccept(OnClientConnected, null); } /// Waits for the upcoming messages from different clients and fires the proper event according to the packet type. /// /// private void WaitForData(IAsyncResult asyncResult) { ClientInfo sendingClient = null; try { //Take the client information from the asynchronous result resulting from the BeginReceive sendingClient = asyncResult.AsyncState as ClientInfo; // If client is disconnected, then throw a socket exception // with the correct error code. if (!IsConnected(sendingClient.WorkerSocket)) { throw new SocketException(ClientDisconnectedErrorCode); } //End the pending receive request sendingClient.WorkerSocket.EndReceive(asyncResult); //Fire the appropriate event FireMessageTypeEvent(sendingClient.ConvertBytesToPacket() as BasePacket); // Begin receiving data from this client sendingClient.WorkerSocket.BeginReceive(sendingClient.Buffer, 0, BasePacket.GetMaxPacketSize(), SocketFlags.None, new AsyncCallback(WaitForData), sendingClient); } catch (SocketException e) { if (e.ErrorCode == ClientDisconnectedErrorCode) { // Close the socket. if (sendingClient.WorkerSocket != null) { sendingClient.WorkerSocket.Close(); sendingClient.WorkerSocket = null; } // Remove it from the hash table. connectedClients.Remove(sendingClient.PlayerId); if (ClientDisconnectedEvent != null) { ClientDisconnectedEvent(this, new ClientDisconnectedEventArgs(sendingClient.PlayerId)); } } } catch (Exception e) { // Begin receiving data from this client sendingClient.WorkerSocket.BeginReceive(sendingClient.Buffer, 0, BasePacket.GetMaxPacketSize(), SocketFlags.None, new AsyncCallback(WaitForData), sendingClient); } } /// /// Broadcasts the input message to all the connected clients /// /// public void BroadcastMessage(BasePacket message) { byte[] bytes = message.ConvertToBytes(); foreach (ClientInfo client in connectedClients.Values) { client.WorkerSocket.BeginSend(bytes, 0, bytes.Length, SocketFlags.None, SendAsync, client); } } /// /// Sends the input message to the client specified by his ID. /// /// /// The message to be sent. /// The id of the client to receive the message. public void SendToClient(BasePacket message, int id) { byte[] bytes = message.ConvertToBytes(); (connectedClients[id] as ClientInfo).WorkerSocket.BeginSend(bytes, 0, bytes.Length, SocketFlags.None, SendAsync, connectedClients[id]); } private void SendAsync(IAsyncResult asyncResult) { ClientInfo currentClient = (ClientInfo)asyncResult.AsyncState; currentClient.WorkerSocket.EndSend(asyncResult); } /// Fires the event depending on the type of received packet /// /// The received packet. void FireMessageTypeEvent(BasePacket packet) { switch (packet.MessageType) { case MessageType.PositionUpdateMessage: if (PositionUpdateMessageEvent != null) { PositionUpdateMessageEvent(this, new PositionUpdateEventArgs(packet as PositionUpdatePacket)); } break; } } } The events fired are handled in a different class, here are the event handling code for the PositionUpdateMessage (Other handlers are irrelevant): private readonly Hashtable onlinePlayers = new Hashtable(); /// /// Constructor that creates a new instance of the GameController class. /// private GameController() { //Start the server server = new Thread(networkManager.StartServer); server.Start(); //Create an event handler for the NewClientEvent of networkManager networkManager.PositionUpdateMessageEvent += OnPositionUpdateMessageReceived; } /// /// this event handler is called when a client asks for movement. /// private void OnPositionUpdateMessageReceived(object sender, PositionUpdateEventArgs e) { Point currentLocation = ((PlayerData)onlinePlayers[e.PositionUpdatePacket.PlayerId]).Position; Point locationRequested = e.PositionUpdatePacket.Position; ((PlayerData)onlinePlayers[e.PositionUpdatePacket.PlayerId]).Position = locationRequested; // Broadcast the new position networkManager.BroadcastMessage(new PositionUpdatePacket { Position = locationRequested, PlayerId = e.PositionUpdatePacket.PlayerId }); }

    Read the article

  • Solving embarassingly parallel problems using Python multiprocessing

    - by gotgenes
    How does one use multiprocessing to tackle embarrassingly parallel problems? Embarassingly parallel problems typically consist of three basic parts: Read input data (from a file, database, tcp connection, etc.). Run calculations on the input data, where each calculation is independent of any other calculation. Write results of calculations (to a file, database, tcp connection, etc.). We can parallelize the program in two dimensions: Part 2 can run on multiple cores, since each calculation is independent; order of processing doesn't matter. Each part can run independently. Part 1 can place data on an input queue, part 2 can pull data off the input queue and put results onto an output queue, and part 3 can pull results off the output queue and write them out. This seems a most basic pattern in concurrent programming, but I am still lost in trying to solve it, so let's write a canonical example to illustrate how this is done using multiprocessing. Here is the example problem: Given a CSV file with rows of integers as input, compute their sums. Separate the problem into three parts, which can all run in parallel: Process the input file into raw data (lists/iterables of integers) Calculate the sums of the data, in parallel Output the sums Below is traditional, single-process bound Python program which solves these three tasks: #!/usr/bin/env python # -*- coding: UTF-8 -*- # basicsums.py """A program that reads integer values from a CSV file and writes out their sums to another CSV file. """ import csv import optparse import sys def make_cli_parser(): """Make the command line interface parser.""" usage = "\n\n".join(["python %prog INPUT_CSV OUTPUT_CSV", __doc__, """ ARGUMENTS: INPUT_CSV: an input CSV file with rows of numbers OUTPUT_CSV: an output file that will contain the sums\ """]) cli_parser = optparse.OptionParser(usage) return cli_parser def parse_input_csv(csvfile): """Parses the input CSV and yields tuples with the index of the row as the first element, and the integers of the row as the second element. The index is zero-index based. :Parameters: - `csvfile`: a `csv.reader` instance """ for i, row in enumerate(csvfile): row = [int(entry) for entry in row] yield i, row def sum_rows(rows): """Yields a tuple with the index of each input list of integers as the first element, and the sum of the list of integers as the second element. The index is zero-index based. :Parameters: - `rows`: an iterable of tuples, with the index of the original row as the first element, and a list of integers as the second element """ for i, row in rows: yield i, sum(row) def write_results(csvfile, results): """Writes a series of results to an outfile, where the first column is the index of the original row of data, and the second column is the result of the calculation. The index is zero-index based. :Parameters: - `csvfile`: a `csv.writer` instance to which to write results - `results`: an iterable of tuples, with the index (zero-based) of the original row as the first element, and the calculated result from that row as the second element """ for result_row in results: csvfile.writerow(result_row) def main(argv): cli_parser = make_cli_parser() opts, args = cli_parser.parse_args(argv) if len(args) != 2: cli_parser.error("Please provide an input file and output file.") infile = open(args[0]) in_csvfile = csv.reader(infile) outfile = open(args[1], 'w') out_csvfile = csv.writer(outfile) # gets an iterable of rows that's not yet evaluated input_rows = parse_input_csv(in_csvfile) # sends the rows iterable to sum_rows() for results iterable, but # still not evaluated result_rows = sum_rows(input_rows) # finally evaluation takes place as a chain in write_results() write_results(out_csvfile, result_rows) infile.close() outfile.close() if __name__ == '__main__': main(sys.argv[1:]) Let's take this program and rewrite it to use multiprocessing to parallelize the three parts outlined above. Below is a skeleton of this new, parallelized program, that needs to be fleshed out to address the parts in the comments: #!/usr/bin/env python # -*- coding: UTF-8 -*- # multiproc_sums.py """A program that reads integer values from a CSV file and writes out their sums to another CSV file, using multiple processes if desired. """ import csv import multiprocessing import optparse import sys NUM_PROCS = multiprocessing.cpu_count() def make_cli_parser(): """Make the command line interface parser.""" usage = "\n\n".join(["python %prog INPUT_CSV OUTPUT_CSV", __doc__, """ ARGUMENTS: INPUT_CSV: an input CSV file with rows of numbers OUTPUT_CSV: an output file that will contain the sums\ """]) cli_parser = optparse.OptionParser(usage) cli_parser.add_option('-n', '--numprocs', type='int', default=NUM_PROCS, help="Number of processes to launch [DEFAULT: %default]") return cli_parser def main(argv): cli_parser = make_cli_parser() opts, args = cli_parser.parse_args(argv) if len(args) != 2: cli_parser.error("Please provide an input file and output file.") infile = open(args[0]) in_csvfile = csv.reader(infile) outfile = open(args[1], 'w') out_csvfile = csv.writer(outfile) # Parse the input file and add the parsed data to a queue for # processing, possibly chunking to decrease communication between # processes. # Process the parsed data as soon as any (chunks) appear on the # queue, using as many processes as allotted by the user # (opts.numprocs); place results on a queue for output. # # Terminate processes when the parser stops putting data in the # input queue. # Write the results to disk as soon as they appear on the output # queue. # Ensure all child processes have terminated. # Clean up files. infile.close() outfile.close() if __name__ == '__main__': main(sys.argv[1:]) These pieces of code, as well as another piece of code that can generate example CSV files for testing purposes, can be found on github. I would appreciate any insight here as to how you concurrency gurus would approach this problem. Here are some questions I had when thinking about this problem. Bonus points for addressing any/all: Should I have child processes for reading in the data and placing it into the queue, or can the main process do this without blocking until all input is read? Likewise, should I have a child process for writing the results out from the processed queue, or can the main process do this without having to wait for all the results? Should I use a processes pool for the sum operations? If yes, what method do I call on the pool to get it to start processing the results coming into the input queue, without blocking the input and output processes, too? apply_async()? map_async()? imap()? imap_unordered()? Suppose we didn't need to siphon off the input and output queues as data entered them, but could wait until all input was parsed and all results were calculated (e.g., because we know all the input and output will fit in system memory). Should we change the algorithm in any way (e.g., not run any processes concurrently with I/O)?

    Read the article

< Previous Page | 118 119 120 121 122 123 124 125 126 127 128 129  | Next Page >