I’m not sure why, but my service keeps going down and I’ll get an email from Virtualmin, then a little bit later, it comes up on it’s own without any intervention.
Where should I check to fix this. It’s happening across all of my websites at different times.
Thanks in advance for any assistance!
It’s an illusion. I’m guessing caused by low memory on your system. The service isn’t really going down (because if it were, it’d probably stay down until you did something about it, and it would be all of your virtual servers rather than random ones, since they all run under one Apache instance)…the status monitor is just taking too long to return and is timing out (which is exactly the same as a failure as far as the monitor is concerned).
So, check memory first.
Then check CPU usage.
You can see both of these in realtime using the top command (Virtualmin also has a memory and CPU usage graph that you can look at). If neither of those looks amiss, we’ll have to look deeper. But, I suspect one or both of them is the culprit, and we’ll need to address it by reducing memory usage on your system or dealing with the runaway process(es).
Are you actually seeing the website go down, or is Virtualmin just telling you it’s down?
I actually disabled the website checks (Status Monitoring) on my system, I found it generating too many false positives
But, one thing to try would be to bump up the timeout for the monitoring – if it was taking too long to respond, even if it was working, that could cause you to get a notification.
I unfortunately don’t have a system available ATM with Status Monitoring enabled, so I can’t tell you exactly where to look.
If you don’t see it, just yell and either I’ll enable that on one of these systems, or perhaps someone else who uses it can provide some insight
Oh, I didn’t make clear that if you only want to use Virtualmin/Webmin to see what’s going on, rather than top, you’ll need to take a look in the Running Processes module to actually find what processes are eating memory/CPU. The graphs just show overall usage.
Thanks. gnome is one that keeps using the most cpu%. I’ve been looking for the past 5 minutes and it jumped from 3% to 21%, but I haven’t seen anything that looks like it’s really taxing the system. rpmq came up and took 80% for a couple of seconds earlier but that was it.
My memory is around 450 of 822.
oh and the websites are going down…
So are you using Gnome for anything? (e.g. is it a local machine that you log on to with a desktop and such?) If not, you’ll want to kill it. Switch your system to boot into runlevel 3 rather than runlevel 5 (runlevel 5 is “everything plus X” while runlevel 3 is “everything, and definitely not X”).
You can alter this in Webmin in the System:Initial System Bootup module. Click on "id" (which stands for "initdefault") and set the Bootup Runlevel to 3 (and uncheck 5).
You’ll then either need to reboot, or use the Webmin System:Bootup and Shutdown module to change the current runlevel to 3 to get rid of X and Gnome.
Obviously, you don’t want to do this if you actually use the Gnome desktop on the system.
yes, i do use gnome because i need the gui to move around with my limited knowledge of centos.
Earlier someone said something about extending the timeout. Can you do that?