This is a really weird problem i am facing... Everything works fine, i love the settings and everything, however, the problems arise when downloading from server.
I mainly use my VPS as temporary storage. So i installed CentOS 5.5 along with Virtualmin. So i transfer files to my server using a custom downloader script (the script terminates after download). So then i download the file from server to my PC using IDM.
I paste the direct link to the file into IDM and click download. Within 5 seconds, my apache is unresponsive, while IDM is downloading the file fine. Its not only me, its happening for everyone. Noone can load ANY pages on the server when the download is running from my or anyone else's computer.
However when i go to port 10000, webmin is loading fine without any problem. Also even if i stop IDM's download, the apache is still unresponsive, so downloads wont resume either. I have to log into ssh and restart httpd in order to get it working again.
Does anyone have any idea why this is happening? I tryed reinstalling httpd, switching to different control panel, even using lighttpd, the problem still persists.
My VPS is 1gbps (shared between my 4 VPS's with almost zero traffic), so its not a bandwidth issue. Also remember that i can access the control panel or any different port without any trouble, just port 80 gives problems.
Also, i only have iptables installed (and i even tryed uninstalling it, didn't help).
Please let me know how can i fix this issue. Thanks :)
Howdy,
Well, what you're experiencing certainly isn't normal :-)
If I had to guess, it's likely either an issue with that particular script, or perhaps an issue of running low on RAM.
I'd start by looking in /var/log/messages, /var/log/httpd/error_log, and $HOME/logs/error_log. Do you see any error messages around the time Apache becomes unresponsive?
What about if you run "dmesg", at the end of that output, is there anything unusual?
Also, how much RAM is available on your system now? You can determine that by running "free -m".
-Eric
What is your MaxClients option set to in your /etc/httpd/conf/httpd.conf file?
-Eric
Howdy,
Well, you really shouldn't be seeing that kind of traffic... that means that you have either 150 or 256 active connections being processed by Apache (depending on if you're using Apache's PreFork or Worker mode... most folks are using the default PreFork).
So if you're using PreFork, that means there's 150 active connections -- that's a lot! It takes a really high traffic site to use up that many connections.
You could of course bump up that 150 MaxClients to a higher number, but I think there's a bigger issue occurring at the moment. I'd suggest reviewing your $HOME/logs/access_log file, and determining where all the traffic is coming from, and what it's accessing.
-Eric
Howdy,
According to the Apache docs, those aren't anything in particular to worry about:
http://wiki.apache.org/httpd/InternalDummyConnection
You may want to scan through all your various Apache logs (ie, for all your domains), and see if you're seeing a high number of external connections in any of them.
-Eric
Howdy,
There's two different MaxClients settings in Apache... one for when Apache is using Prefork, the other when it's using the Worker mode.
Prefork is the default, so I'd suggest starting with that one.
Also, make sure that there isn't a ServerLimit setting somewhere else in your Apache config.
Lastly -- if you set Apache to allow 2000 MaxClients, and each Apache process is roughly, say, 20MB, that means that you'd require roughly 40GB of RAM in your server :-)
I'd recommend making sure that you have enough RAM to handle whatever MaxClients setting you pick.
-Eric