httpd-DSSL processes not closing, creating HUGE server loads

kennyp

Registered
May 29, 2003
4
0
151
Hi,
Im having a very wierd problem it seems every ten minutes or so, sometimes a bit longer i get these processes, either root or nobody as user that do not close and eventualy suck the life out of the cpus.

Ive tried every fix posted in these forums and so far no luck at fixing it. I have 5, dual zeon machines, each running a gig of memory. 4 of my machines have this problem in different degrees.

I cant match the processes to a site in Apache usage on WHM because they are no longer listed in apache processes although i do see then in the process list and when i call up top via ssh . I can sit and watch top and see thet process start and almost set my clock to thet server load rising. If I kill the process, usually a nobody one with a different pid one comes along and the only temp solution i have found is rebooting apache and then things are fine till the next http-DSSL process comes along and it doesnt seem to afftect anything till its been running for a few minutes which i dont think is normal.

Ive been running my servers on my own for 3 years and although self taught seem to have a good handle on thngs for the most part. This is the first time I have seen anything like this .

If i helps the machine that is the worst its the only I have that a master/slave (its thet slave) configturtaion or I should say the new clustering configuration which I suspect may be part of the problem. Dont know why they changed it, master/slave always worked fine for me in the past .

One last thing, if I kill the root process then kill the nobody one that always seems to show up, apache seems to crash as I cant access any webpages, although if i just reboot apache, processes disappear and server load drops to normal levels

First thought corupted apache, i have reinstalled numerous times as well as reinstalling exim. On the machine with the most problems Im running redhat 9 with cpanel 9.1 stable 93.

If anyone knows the source of my protblem and a fix I would be etertnally grateful.

Thanks in advance

Kenneth Poirier
ByyteWorks Web Services
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
Hi,

I'm having a similar problem althought I'm not sure
it's the same source causing it.

Every now and then the cpu load goes up to high
1,5-2,0 and the after a little, little while it goes down to
normal - 0,2.

2 days in a row my server has crashed and the NOC
had to reboot it. I suspect it's the server load going
above what it can take but I have never seen it live.

I can't read the logs to see why this is happening.

Before these 2 days everyhting was fine and I can't think
of anything changed during this period.

When I', looking live at the server I have onlu seen
it go up to 0,5 or so and then its: httpd-DSSL processes
running.
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
Well, I upgraded to the release version of CP
and ever since it has not crashed.
 
Jul 14, 2003
17
0
151
We have the same problem.

RHEL 3
cPanel 9.2.0-E21

This looks like a Cpanel bug...like Cpanel is not correctly communicating with Apache to terminate the sessions.

Having to kill these manually is getting VERY old. :mad:
 

netlook

Well-Known Member
Mar 25, 2004
334
0
166
We aslo :(

WHM 9.1.0 cPanel 9.1.0-R85
Fedora - WHM X v2.1.2

Any idea to solve it?
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
I have not had any crash and the times I look I can't
see any real high load anymore.

What I did was to upgrade to the latest Release cPanel,
stop the daily backups for one day and then back on.

Worked for me, maybe a coinsident.
 

netlook

Well-Known Member
Mar 25, 2004
334
0
166
To which version did you upgraded cPanel? And please tell me how did you that without any problems ;)
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
Now I have:

WHM 9.2.0 cPanel 9.2.0-R20
RedHat Enterprise 3 - WHM X v2.1.2

this is the second time I updated since this issue
Before when the issue was fixed I had another version
9.1xxxx something.

I just click on update to latest cpanel :)
always worked fine with me.
 

netlook

Well-Known Member
Mar 25, 2004
334
0
166
/scrpits/upcp worked for me well :)

but...

didn't solve high server loads problem. I think it shoul be memory problem - now I have only 512MB and almost 200 accounts, so it might be to low, because most of them have phpBB installed and use lot of mysql.

and another funny thing - I've translated my cPanel into my country language, and I forgot to backup copy of this language file, so after updating cPanel you know... silly, silly...
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
oh my god :), sorry about that.
You know rvskin.com has alot of languages and is a really
great skin/themes with much better "select what is included in the account" kind of stuff. Really great.
 

mydomain

Well-Known Member
Aug 10, 2003
138
0
166
Still getting 1 or 2 httpd processes using all the CPU up as originally reported on RHEL with 9.2.0-R24.

Dont know what to do about this, tried many things
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
Some say it's because of Tomcat and some say it's because of a SSL setting. Who knows...

I don't think it's either of it...
STrange no expert can answer this
 

yaax

Well-Known Member
Jun 15, 2003
67
0
156
Bug in Cpanel 9.2 R24 on RHEL 3

It seems that it is really bug in latest update of Cpanel on Rhel 3 causing me huge server loads on server with RHEL 3.

I see hundreds of Apache httpd processes running and only when I restart httpd they are working ok.

Does anyone know how to fix this? or how to downgrade cpanel to latest working version without this bug?
 

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
two days ago and today these processes nearly killed
my server again. I'm happy I did install the
PRM from rfxnetworks. Now the server was only in critical state for about 3-5 minutes before the processes was killed.

But sure something is really wrong here.
BUt this is only on my celeron server not on my xeon server
with same configuration.
 

yaax

Well-Known Member
Jun 15, 2003
67
0
156
I have solved it buy addition of httpd restart each 2 minutes in crontab.

Before this I tried httpd restart each 5 minutes and still has problems with loads jump up to 100, only after httpd restarts each 2 minutes I see server is quite stable.

It is only on my server with RHEL 3 , my another servers with RH 9 and Fedora are working ok with this version CPanel.

Still this a bug and I already opened it in bugzilla of cpanel:
http://bugzilla.cpanel.net/show_bug.cgi?id=464

I am not sure this is a HUGE bug request, but still....
 
Last edited:

jeroman8

Well-Known Member
Mar 14, 2003
410
0
166
You restart apache every 2 minutes ??

That means your websites can't be reached
during 10 seconds every 2 minutes or so.