This is wierd. Apache quits on me all of a sudden. Logging into ensim webppliance verifies this. I do NOT restart apache. After a while, apache is back up. Its been like this for the whole of today. Downtime is between 1-15 minutes per incidence. And uotime is between 1-5 hours.

What is wrong? What can I do?

This is just an educated guess, but... I would check if there is a cron job (under root) to check if the apache process is active. My guess is that there is something like that and it's restarting apache.

su - root
crontab -l

That will switch to root, then list the crontab.

Check the main apache error log to see if there is any entries related to the "crash".

Frank

Whats the server load like? Could be a crontab problem like mentioned above. Or maybe because of the server load, apache just takes time to respond to a http request!

Sounds like a run away process. Watch your server load and look for a process taking up a lot of resource. Maybe a infinte loop in a PHP file some one has loaded on the server.

no crontab for root

how can i check the server load? and running processes?

thanks for any help....newbie true to my username :D

Which version of Apache are you running? It's entirely possible that someone is chunking you to death. You might want to look into installing mod_blowchunks or upgrading to 1.2.36 ASAP.

If this is Ensim LS 3.1 I've been having problems with apache too, though they seem to have stopped now. Apache froze a few times for no apparent reason and the only way to get it to restart was by rebooting the server. Couldn't find anything in the error logs even suggesting apache had frozen, let allong what had caused it. It may just be some shifty coding in Ensim, which wouldn't suprise me at all as I've had more problems than this one since updating to LS 3.1. Needless to say I'm switching to H-Sphere asap.

Originally posted by linuxnewbie
no crontab for root

how can i check the server load? and running processes?

thanks for any help....newbie true to my username :D
Login using SSH or telnet and run: top
press q to quit
exit to logoff
You can use Sysinfo (http://phpsysinfo.sourceforge.net/) it is a good PHP based program, no install required.

"ps aux" will give you process list
Also if apache is from rpm package or other binary, then you should compile it by yourself. (http://httpd.apache.org for docs)

thanks for all the help.....im running apache 1.3.19 which came with ensim 3.0
i have applied all the patched on ensim website.

typing ps aux

gave me a list and cpu and mem usage was very low for all except one which showed:

CPU: 49% MEM: 75

it had username admin16 next to it and to the far right it showed /usr/sbin/perl bf3

So i think its this one thats the problem, how do i decipher the results from ps aux...can someone take a moment to explain?

thanks very much

Originally posted by Miha
"ps aux" will give you process list
Also if apache is from rpm package or other binary, then you should compile it by yourself. (http://httpd.apache.org for docs)

Ensim Webppliance users should NOT attempt to compile Apache on an Ensim box.... (unless that really, really know what they're doing ...)

Ensim uses a RPM for apache that is somewhat "non-standard" apache configuration. Compiling apache _will_ break the Webppliance user interface (unless you know the ends and outs of how Ensim has configured Apache).

Unfortunatley, Ensim doesn't make the src RPM's available, that I know of....

Originally posted by linuxnewbie
thanks for all the help.....im running apache 1.3.19 which came with ensim 3.0
i have applied all the patched on ensim website.

typing ps aux

gave me a list and cpu and mem usage was very low for all except one which showed:

CPU: 49% MEM: 75

it had username admin16 next to it and to the far right it showed /usr/sbin/perl bf3

So i think its this one thats the problem, how do i decipher the results from ps aux...can someone take a moment to explain?

thanks very much

I'm not sure how Ensim works, but I hate it and I've made small HOWTO on removing ensim (http://re.4-tz.net).
About 'ps aux' - in standard linux/BSD box if you're logged in shell as simple user (not root) it will give you the list of all processes on the machine unless the 'ps' is not patched! Some unix admins likes to patch 'ps' so it will not show all processes to user, but root of course can view all processes. So one possibility is that ensim also patched 'ps' for simple users and as you said you were logged in as admin16, which is not the same as root. What you should do to view full list of processes is you need to login as root (no admin, but root) and then 'ps aux'
Here is example of 'ps aux' on my home PC right now (the list is not full for security reasons):

named 702 0.0 0.5 10568 2592 ? S 06:30 0:00 named -u named
named 704 0.0 0.5 10568 2592 ? S 06:30 0:00 named -u named
named 705 0.0 0.5 10568 2592 ? S 06:30 0:00 named -u named
named 706 0.0 0.5 10568 2592 ? S 06:30 0:00 named -u named
named 707 0.0 0.5 10568 2592 ? S 06:30 0:00 named -u named
root 727 0.0 0.1 2672 1024 ? S 06:30 0:00 /usr/sbin/sshd
root 762 0.0 0.1 2248 1008 ? S 06:30 0:00 /bin/sh /usr/bin/safe_mysqld -
mysql 802 0.0 0.8 26824 4340 ? S 06:30 0:00 /usr/libexec/mysqld --defaults
mysql 805 0.0 0.8 26824 4340 ? S 06:30 0:00 /usr/libexec/mysqld --defaults
mysql 806 0.0 0.8 26824 4340 ? S 06:30 0:00 /usr/libexec/mysqld --defaults
mysql 816 0.0 0.8 26824 4340 ? S 06:30 0:00 /usr/libexec/mysqld --defaults
root 818 0.0 0.2 5224 1384 ? S 06:30 0:02 sendmail: accepting connection
root 837 0.0 0.0 1444 480 ? S 06:30 0:02 gpm -t netmouse -m /dev/mouse
root 855 0.0 0.1 1592 656 ? S 06:30 0:00 crond

There are 106 processes on my machine right now, most of them 'sleeping'. I'm under user 'miha', but not under user root, I can view processes because I haven't patched 'ps'.

Hope it helps.

hi, i was logged in as root. i got a huge list just like yours but the only listing of concern was the one which showed :

CPU: 49% MEM: 75

originally appeared something like this:

admin16 855 49.0 75.0 1592 656 ? S 06:30 0:00 /usr/bin/perl bf3

thats what i meant by "had admin16 next to the CPU:49% and mem:75%" I was ofcoursse logged in as root.

I was hoping someone could explain the listing... I understand the 1st column is the owner of the process, 2nd is the pid, 3rd the cpu usage in % and 4th the memeory usage in %. I dont know about the other columns, if you could explain that, id be most grateful.

also today morning i saw a process with 0.00 cpu and 49memory for an apache process...is this wierd...how can i stop this from happening

Let's take one of processes on my home pc:

root 4466 0.0 0.4 4044 2280 ? S Jun10 0:22 snmpd -a

The first column ('root') is showing the username that running process now.
The second column ('4466') is PID (Process ID).
Third one (0.0) is how much % the process took from your processor, your processor has 100% of total proc, and process tokk 0.0.
Fourth (0.4) is how much memory it uses. You have 100% of memory total, 0.4 is 0.4% of 100%.
Fifth (4044) shows PID that started this process.
Sixth (2280) is also related to the same as fifth.
Seventh (?) is TTY that process is using, and if it is '?' then probably there is no TTY for this process, or system can't guess.
Eight (S) is showing current state of process. 'S' means 'sleep'
Ninth (Jun10) is the date when process started.
Tenth (0:22) is CPU time which is related to server load
And the last (snmpd -a) is what command is it.

Also you can 'ps auxwwww' which will show you more detailed list with all processes in the chain.

Hope now you can understand it better.

Miha

thanks a whole lot! :)