Re: postgres/shm problem

Re: postgres/shm problem

[manfred]

Do you run powertweak?

there is a new parameter

	/proc/sys/kernel/shmall

and some powertweak versions set it to 0.

--
	Manfred
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: postgres/shm problem

[Narancs 1]

On Mon, 8 Jan 2001, Nils Philippsen wrote:

> 
> #define SHMALL (SHMMAX/PAGE_SIZE*(SHMMNI/16)) /* max shm system wide (pages)
> */
> On my machine here, it is 2097152. It should be the same on any Intel IA32
ok thanks for it! I donwanna reboot my computer ;-)

10x4all
Narancs v1

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: postgres/shm problem

[Alan Cox]

> Is there any other solution than recompiling programs?
> all of the program which need shm/ipc will turn blue?

Congratulations you are running an old version of powertweak. Remvoe that
and it should recover (its setting shm limits to 0)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: postgres/shm problem

[Nils Philippsen]

On Mon, 8 Jan 2001, Narancs 1 wrote:

> On Mon, 8 Jan 2001, Nils Philippsen wrote:
>
>
> > /proc/sys/kernel/shmall to "0" (that is the maximum number of SHM
> > segments).
> yes, powertweak made it wrong. what is the good value for it?

according to /usr/src/linux/include/linux/shm.h:

#define SHMALL (SHMMAX/PAGE_SIZE*(SHMMNI/16)) /* max shm system wide (pages)
*/

On my machine here, it is 2097152. It should be the same on any Intel IA32
compatible machine. To be safe, remove the value from /etc/sysctl.conf and
reboot the machine ;-).

Nils
-- 
 Nils Philippsen / Berliner Straße 39 / D-71229 Leonberg // +49.7152.209647
nils@wombat.dialup.fht-esslingen.de / nils@fht-esslingen.de / nils@redhat.de
   The use of COBOL cripples the mind; its teaching should, therefore, be
   regarded as a criminal offence.                  -- Edsger W. Dijkstra


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: postgres/shm problem

[Narancs 1]

On Mon, 8 Jan 2001, Nils Philippsen wrote:


> /proc/sys/kernel/shmall to "0" (that is the maximum number of SHM segments).
yes, powertweak made it wrong.
what is the good value for it?

thx
Narancs v1

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

Re: postgres/shm problem

[Nils Philippsen]

Hi,

On Mon, 8 Jan 2001, Narancs 1 wrote:

> Lots of things may have changed in 2.4.0-prer., because nor cdrecord
> neither postgressql do not run because of system v ipc / shm problem.
> After recompiling cdrecord with the new header files, started working.
> I'm just recompiling postgres too.
>
> /var/log/postgresql.log:
>
> 010102.10:45:09.448   [627] IpcMemoryCreate: shmget failed (No space left
> on device) key=5432010, size=144, permission=700
> This type of error is usually caused by an improper
> shared memory or System V IPC semaphore configuration.

I had the same problem on my machine and it stemmed from accidentally setting
/proc/sys/kernel/shmall to "0" (that is the maximum number of SHM segments).
In my cas I used some /psoc/sys tweaking utility at a time where this value
was unknown (other kernel) and that utility then set kernel.shmall in
/etc/sysctl.conf to 0. You might check that.

Nils
-- 
 Nils Philippsen / Berliner Straße 39 / D-71229 Leonberg // +49.7152.209647
nils@wombat.dialup.fht-esslingen.de / nils@fht-esslingen.de / nils@redhat.de
   The use of COBOL cripples the mind; its teaching should, therefore, be
   regarded as a criminal offence.                  -- Edsger W. Dijkstra

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/

postgres/shm problem

[Narancs 1]

Dear Kernel developers!

Lots of things may have changed in 2.4.0-prer., because nor cdrecord
neither postgressql do not run because of system v ipc / shm problem.
After recompiling cdrecord with the new header files, started working.
I'm just recompiling postgres too.

/var/log/postgresql.log:

010102.10:45:09.448   [627] IpcMemoryCreate: shmget failed (No space left
on device) key=5432010, size=144, permission=700
This type of error is usually caused by an improper
shared memory or System V IPC semaphore configuration.
For more information, see the FAQ and platform-specific
FAQ's in the source directory pgsql/doc or on our
web site at http://www.postgresql.org.
010102.10:45:09.481   [627] IpcMemoryIdGet: shmget failed (No such file or
directory) key=5432010, size=144, permission=0
010102.10:45:09.481   [627] IpcMemoryAttach: shmat failed (Invalid
argument) id=-2
010102.10:45:09.482   [627] FATAL 1:  AttachSLockMemory: could not attach
segment
010105.08:52:27.995 [12675] IpcMemoryCreate: shmget failed (No space left
on device) key=5432010, size=144, permission=700

cdrecord says the same no space left on device mesg.

Is there any other solution than recompiling programs?
all of the program which need shm/ipc will turn blue?

kernel 2.4.0-prerelease
postgresql                 7.0.3-2 postgresql                 7.0.3-2 
cdrecord 1.9


thx
Narancs v1


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/