CentOS 5 Networking :: Configuring For Large Number Of Tcp Connections?
Jan 24, 2011
i've got a select based application that wants to support a large number of mostly idle connections. the code is java and works on windows, suse enterprise linux, mac os x. it does not work on centos 5.5 (32-bit, 2.6.18 kernel, 1G of memory).
i've read and followed the directions in various articles about tuning linux for large numbers of connections (including the C10K problem), and gotten the number of sockets up to 3200.
these didn't make any apparent difference:
[URL]
on windows, i can get up to around 78,000.
on suse enterprise linux (a few years ago), i got up to 90,000. that's where i got bored and stopped.
on my mac laptop with os x (snow leopard), i got up to 10,500.
i have used ulimit -n 10240
my current goal is 10k sockets.
the test is that i'm opening one socket at a time until it fails. when it fails, many of the sockets which have already been opened also fail, in one giant cascade. sounds like a buffer / memory problem.
each group of 64 sockets gets a thread to manage select calls for them. thus i'm only using around 61 threads total when it fails.
I am attempting to upgrade a system from 4.7 to 5.2 using a (now) DVD drive attached to the onboard IDE. Originally I had tried using a remote NFS image and a USB stick but I thought maybe there was a problem with the image. I can get up to the point of the installation of selecting the keyboard for the system and then it freezes and never goes any further. It doesn't appear to be a kernel panic since I can still switch between consoles.
I've got an MSI K9NGM2-FID with 14 drives in it. It serves as a file server for our backup server. It's got a secondary 4 port Silicon Image SII 3114 SATA card using the sata_sil module, and an old IDE Promise FastTrak TX2000. Technically I could have 16 drives but the 750W PS is walking the fine line on tripping it's self-breaker with the 14 drives and 7 fans. I would like to NOT have to disconnect all of this to do the upgrade.
I thought maybe that running the install using the "noprobe" option would help so it didn't detect and load the modules for the Silicon Image or the Promise cards and detect all of the drives but it still gets stuck on the step after selecting the keyboard. The installation info console and the dmesg console don't really provide any useful information. The installation console says:
INFO : moving (1) to step welcome INFO : moving (1) to step language INFO : moving (1) to step keyboard INFO : moving (1) to step findrootparts
And the last lines of the dmesg console says:
<6>device-mapper: multipath: version 1.0.5 loaded <6>device-mapper: multipath round-robin: version 1.0.0 loaded <6>device-mapper: multipath emc: version 0.0.3 loaded
Is there a hidden "debug" option that will turn on a lot of extra logging?
How to number of connections for a single ip on port 80 to CentOS 5.5 with iptables? connlimit did not work on CentOS and nginx does not provide a module for that
How do I find the maximum number of concurrent connections (in any state)? I'm running RHEL5 2.6.18-194.26.1.el5. Also, does tcp auto tune affect the number of concurrent connections or is it mostly used for dynamic buffer size allocation?
i am too new for linux..I installed Ubuntu 10.10 desktop edition from windows.When i am trying to configure network connections i am facing problems
Code: udo pppoeconf no working ethernet card could be found.If you do have an interface card which was not autodetected so far,you probably wish to load the driver manually using the modconf utility.Run modconf now? /usr/sbin/pppoeconf: 523: modconf: not found Ifconfig
I've a directory containing around 2.8 lacs of files. I want to move them to another directory.If I use cp or mv then I get an error 'argument list too long'. If I write a script like
for file in ls *; do cp {source} to {destination} done
then because of ls command , its performance degrades.How can I do this?
I am facing problem in copying a large number of file 18 lakh (18,000,000) files from my personal hardisk to another hardisk each file is very small and size of folder is around 3.95 GB copying files using copy given by Windows is frustrating and I am not even able to compress file its giving me error that its not readable.And problem is I am not able to open this drive in Linux it showing me error there saying do diskchk in Windows and Windows disk check is also not able to repair this drive and goes into some mode unsolvable.Is there any way to open disk with error to open in Windows and if not any way I can copy data faster?ERROR: Disk labled EDU is corrupt go to windows and chkdsk /f there and reboot into window 2 times.
I have a large number of log files, on a linux box, I need to cleanse sensitive data from before sending to a third party. I have used the below script on previous occasions to perform this task, and it has worked brilliantly (script was built with some help from here :-)
However, now one of our departments has sent me a CLIENT_FILE.txt with 425000+ variables! I think I may have hit some internal limit. I have tried splitting the client file into 4 with around 100000 variables in each, this still doesn't work. I'm loathe to keep splitting though as I have 20 directories with up to 190 files in each directory to run through. The more client files I make, the more passes I have to do.
I understand that chroot is usually used to provide security, however, for my issue, security is a big don't care. I am very new to using chroot and don't fully understand how the chroot'd env works.
problem: Trying to use a vendor supplied cross compile environment. The environment runs as a chroot'd env and works just fine. I have a large number of additional modules that I wish to compile in the chroot'd environment. FYI, these modules are also (succesfully) compiled for other targets not using chroot'd env's. Copying the source files into the the chroot environment is not an option (don't have hours to wait for copies to finish and it would break the make system). Having them live in the environment is also not an option (the chroot build is a tiny part of the build process and we cannot revamp our entire source tree to accommodate it).
I am looking for a way to have the compiler in the chroot'd env have access to a path that is outside of the env and typically higher up in the same path that holds the chroot'd env. I have tried soft links (they don't work as expected). Hard links only work for single files and there are 10's of thousands of files that would need to be linked. I am not sure how I would go about exporting the additional files and then mounting the exported files in the chroot'd env (or if that would even work).
I have a 50000 line(ish) set of records in a file. I have another file where I have filtered out all the line numbers for those which have an error of various types. e.g column count, field type etc. I want to get all those lines into a separate file so I can sanitise them. There are abt 3-4000 of them.
How can I access those lines which I want to isolate into a single file? I have all the usual linux stuff available and a bit of understanding of regexps.
I need to limit the number of ssh connections a user has. All the users are using tunnel only so their shell is set to /sbin/nologin The logins do not open a shell they just create the tunnel so /etc/security/limits.conf has no effect on them at all.
I tried setting 'MaxSessions 1' in sshd_config but either that doesn't not do what I expect it to or it plain does not work as even with a normal user I was able to open an unlimited number of sessions. I need a good secure way to limit each user to 1 ssh session without them having a shell but Im unable to find a solution.
I used 9.04 for months and it work fine before restarting my PC. After I restarted my PC, the memory consumption takes up to 4.2 GB after login. However, I cannot find any process that consume such large number of memory.
I'm running 64-bit 10.04, upgraded from 9.10. The problem I am experiencing is that any user accounts aside from my main account are problematic. This includes any accounts I add, as well as the GDM guest session.The specific problems that I have thus far experienced are as follows:
1. The desktop loads often improperly. In the latest instance of this the graphics on the right side top panel were randomly chopped-up, leaving parts of my clock on either side of the volume control, among other things. 2. If I make ANY customizations to the desktop at all, the desktop takes nearly a full minute to load on log-in. 3. Flash videos don't work properly on Firefox. Sometimes they only play after refreshing a page, often they will not load at all. Also, attempting to load or play a flash video will sometimes causes Gnome or Firefox to crash. 4. (And this is the one that REALLY has me stumped) Whenever I log into my main accountant after logging out of another account, the IBus control appears in my system tray.However, when I open the IBus preferences the associated check box is (and has always remained) unchecked.Not sure where to go with this one. More than anything, the IBus bug makes me unsure of where to even begin looking for the problem.
This problem is not exclusive to Ubuntu, I've experienced it in Windows and OSX as well, but it seems that almost every time I transfer a large number of files (i.e. my music collection) between my desktop computer and laptop via my external hard drive, I end up losing files for no reason. I usually don't notice the files are missing until later on, because I am never informed of any data loss. Now, every time I make a large transfer of files, I just do it two or three times to ensure that I don't lose any files.
We recovered a large number of files from a HD I messed up. I am attempting to move large numbers of files of a type e.g. .txt .jpg , into a folder by type to more easily sort through them.
Here are the commands I have mainly been trying with various edits:
Code:
Code:
So far the most common complaint I have gotten "missing arguments to execdir".
I'm trying to extract the sender id from a fairly large number of files and am having trouble assigning variables from a file. Here is what I have so far, (which is fairly kludgy I know, but it's been some years since I've done any scripting or programming, and I find that I have lost the knack to a large degree).
I have Ubuntu 9.04 and I would like to know what is the command to find out the number of connections per user (i.e. ssh) and also what syslog module do I need to modify and how to enable it in order to have it logged, thing is I need to generate this event in order to see it in RSA enVision, I've tried with "who" but nothing is displayed, do I need to program something besides that? or what can I do? to have no. of connections and have that logged so I can see the event in enVision.
if i want user should`t have more than 20 sftp connections to a server,is there any way we can limit no.of connections to a particular user on the server using ssh configuration
I have a Intel WIreless 3945 card, wit ipw3945 installed on Dell E1505. I use NetworkManager to function and everything is ok. I can connect to wireless networks that have no security, or WEP security, but I just can't connect to networks with WPA or WPA2 securty. What can be my malfunction?
I have on my windows machine several hundred files that are a format of .nc .ncs for a CNC machine. I need to convert them to txt which is something as easy as opening in notepad and then saving as .txt but there are so many that this kind of action would take way too long.
The reason I am writing the linuxquestions is because I would feel more comfortable in loading a live CD and using some sort of terminal command to do this than I would to download one of the many "freeware" type programs I have found for windows (even more so since I have had a root kit before and had to start all the way over to get rid of it).
I need to know:
1. Is this possible to do with the terminal without super advanced knowledge.
2. Can one please point me in the right direction; something to read or an example
I want to move files from a $SOURCEDIR to a $DESTBASE/$DESTDIR. Under $DESTBASE there are many directories, and I need to test beforehand if a file from $SOURCEDIR already exists in any of them.
This is obviously extremely slow, and the real use case involves dozens of dirs and thousands of files. Creating a temporary "index" file for the find command (instead of running it every iteration) speeds it up a little, but it's still very clumsy.
Recently my motherboard went out, I replaced it and booted centos 5. All went well except sound, video tuners and netoworking. The sound I re-configured, and have gotten tuners working but I am running into trouble with the new network card. The networking setup is;
DSL wireless router/modem IP 192.168.0.1 connected to XP computer through WIFI IP 192.168.0.3. all set to subnetmask 255.255.255.0 and with DNS servers 192.168.0.1 and 205.171.3.25 the XP computer is the only one that supports my wifi card. On wired LAN I share that connection to 4 devices. 3/4 work well, my cent os 5 box will not connect. here are the lan settings:
win xp: IP 192.168.0.60 255.255.255.0 mac 8.6: IP 192.168.0.34 255.255.255.0 blueray: IP 192.168.0.65 255.255.255.0
I have ndiswrapper installed, and have successfully installed the drivers, I have also used nm-applet (and x forwarding) to configure wireless connections. My problem is that I want to reboot and unplug my wired connection and use just my wireless connection, but when I reboot it doesn't seem to connect to my network.
is there a command line tool or an independent tool that i can launch forwarding x I can use to search for and connect to wifi networks that would also have the ability to save configuration so that when I reboot the system automatically connects to my wifi?
except is there is a way to enhance mod_limitipconn.c to ensure that apart from restricting one connection allowed from a given IP, also set so that an IP can only connect on every set interval ?e.g.restrict the number of connections from a given source IP to say once every 5 minutes or so?if not mod_limitipconn.c, any other mechanism to do the expected result?
