topic Re: Oracle versus Apache in Operating System - HP-UX

Oracle versus Apache

jpcast_real — Sat, 09 Apr 2005 07:55:50 GMT

Hello Gurus,

I have found a problem between my Oracle 9i and the default apache web server I have installed in the system . If oracle is running the apache can not be started and if apache is running oracle can not be started .

This is the error log of apache startup when oracle is running.
[Sat Apr 09 14:29:30 2005] [notice] LDAP: Built with OpenLDAP LDAP SDK
[Sat Apr 09 14:29:30 2005] [notice] LDAP: SSL support unavailable
[Sat Apr 09 14:29:30 2005] [crit] (28)No space left on device: mod_rewrite: could not
create rewrite_log_lock
Configuration Failed

This is the error log from oracle when apache is running:

SQL> Connected to an idle instance.
SQL> ORA-00600: internal error code, arguments: [OSDEP_INTERNAL], [], [], [], [], [], [], []
ORA-27302: failure occurred at: skgpwreset1
ORA-27303: additional information: invalid shared ctx
ORA-27146: post/wait initialization failed
ORA-27300: OS system dependent operation:semget failed with status: 28
ORA-27301: OS failure message: No space left on device
ORA-27302: failure occurred at: sskgpsemsper

Whith the ipcs commnad I see the areas of memory used by the application ... This problem is happening in the two new rp4440 systems , however in the old rp5470 everything works fine.

Thanks in advance

Re: Oracle versus Apache

harry d brown jr — Sat, 09 Apr 2005 12:02:20 GMT

It appears you have a lack of disk space somewhere:

No space left on device

live free or die
harry d brown jr

Re: Oracle versus Apache

Bill Hassell — Sat, 09 Apr 2005 18:56:02 GMT

Take a look at bdf output. You're going to need several dozen megs of free space on /var as an example. In the Oracle error messages (cyrptically hidden) is the OS error code 28. Whenever you see OS error codes (or similarly worded messages), these are errno values, a classic error number in Unix. From /usr/include/sys/errno.h:

define ENOSPC 28 /* No space left on device */

from man 2 errno:
[ENOSPC] No space left on device. During a write() to an ordinary file, there is no free space left on the device; or no space in system table during msgget(), semget(), or semop() while SEM_UNDO flag is set.

Re: Oracle versus Apache

Julio Yamawaki — Sat, 09 Apr 2005 20:43:57 GMT

Hi,

This can happen if your init.ora - processes parameter is set
too high, but the kernel is not configured to handle this.
In our case, the message, ' operation:semget failed' indicates
the call to get the required number of semaphores has failed.

Suggestion:
Increase your kernel parameters , specifically the semaphore
related parameters such as SEMMNI and SEMMNS.

OR

Decrease the processes value in the init.ora and try restarting
the instance.

Re: Oracle versus Apache

Florian Heigl (new acc) — Sun, 10 Apr 2005 07:30:44 GMT

You're definitely running out of shared memory or shared memory segments (I'd point to that)

Re: Oracle versus Apache

jpcast_real — Mon, 11 Apr 2005 02:10:23 GMT

I also think that I am out of memory segments
. This is my bdf:

Aramis:/root> bdf
Filesystem kbytes used avail %used Mounted on
/dev/vg00/lvol3 819200 262240 552728 32% /
/dev/vg00/lvol1 497584 44216 403608 10% /stand
/dev/vg00/lvol8 8708096 922168 7725936 11% /var
/dev/vg00/lvol7 2564096 1359848 1194904 53% /usr
/dev/vg00/lvol6 2048000 1439408 603864 70% /tmp
/dev/vg00/lvol5 5120000 2259520 2838200 44% /opt
/dev/vg00/lvol4 516096 443320 72352 86% /home
/dev/vgtemp/lvol3 12288000 2058298 9590402 18% /tmp-mad
/dev/vgpkgdb/503 5120000 1326857 3557859 27% /mad503
/dev/vgora/data 10240000 5705871 4250787 57% /mad503/madora
/dev/vgredo/redo 2560000 631498 1807977 26% /mad503/redo
/dev/vgora/ora 5120000 3137500 1858642 63% /ora

Can anyone tell me how to order the shared memory areas in my servers..

Re: Oracle versus Apache

jpcast_real — Mon, 11 Apr 2005 03:34:16 GMT

Problem solved.

Increasing the shared memory size and semaphores kernel parameters the problem was solved.

Thanks for your help