HP Apache 2.0.48 error on hpux 11.00/11.11i

[Shivkumar]

Dear Sirs,

We are using the HP Apache version:-
Server version: Apache/2.0.48 HP-UX_Apache-based_Web_Server
Server built: Jan 17 2004 02:06:21

Sometime during restarting apache process comes up but we don't access the page. We saw the error in the error
log as below:-

"Thu Dec 15 20:18:07 2005] [error] Cannot allocate shared memory: (17)File exists"


I deleted a file called "ssl_scache" under the ../../../logs directory and it came up successfully and was able to
access the page.

Can anyone suggest the root cause ?

Thanks,
Shiv

[Henk Geurts · Accepted solution]

if i would have this problem and restarting did not work i would check for shared memory segments still in use by the application (ipcs |grep )
and get rid of them(ipcrm -m
Then restart would work...

[Shivkumar]

Henk;

Could you please eloborate your explanation in detail ?

Regards,
Shiv

[Shivkumar]

Is HP Apache 2.0.48 has any shared memory issue ? Should we upgrade to newer version of HP Apache to fix this issue ?

[Sameer_Nirmal]

Hi,

As the error indiactes "File Exist" which does allow apache process to allocate shared memory. There are two more "left over" files "httpd.pid" & "apache_runtime_status" may create "File exists" errors besides "ssl_scache" file.

This usually happens becuase of abnormal shutdown or crash of the apache server process.

Refer following docs for fix.

http://www2.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000079993969

http://www2.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000074247281

[Sameer_Nirmal]

Oops!!!

>As the error indiactes "File Exist" which >does allow apache process to allocate >shared memory

Should be *read* as

As the error indiactes "File Exist" which does not allow apache process to allocate shared memory.

[Steven E. Protter]

Shalom Shivkumar,

HP Apache 2.0.48 has six releases since it dealing with performance, bugs, security, a lot of issues.

I try and keep my apache servers current to avoide these issues.

Your system may have an issue with the kernel settings shmmax and shmseg.

I would consider increasing these parameters if glance shows some kind of shortage. I would also upgrade the apache binaries as well.

The root cause is probably a either a bug, shared memory shortage or a combination of the two.

SEP

[Arunvijai_4]

Hi Shiv, You can try to install and upgrade to HP-UX Web Server Suite v.2.14, since it is based on Apache version 2.0.55, which address many bug fixes and security issues. You can get more information at,.

http://h20293.www2.hp.com/cgi-bin/swdepot_parser.cgi/cgi/displayProductInfo.pl?productNumber=HPUXWSSUITE

-Arun

[Henk Geurts]

hi Shiv.
here it is:
ipcs displays information in short format for the message queues, shared memory segments, and semaphores that are currently active in the system.

So when an application stops abnormally, there a change not all queues,segments and semaphores are cleaned up.
This might cause errors on restart ,the kind you discribe..
Therefor cleaning up can be quite useful (this only works when the application runs with a non-root userid with no other sessions of the application running)

#!/usr/bin/ksh
for i in `ipcs -m | grep | awk '{print $2}'`
do
ipcrm -m $i
done

for i in `ipcs -s | grep | awk '{print $2}'`
do
ipcrm -s $i
done


Regards.