BUG #3495: PostgreSQL does not use POSIX SHM or SEM in place of SYSV equivalents - Mailing list pgsql-bugs

From Roger Leigh
Subject BUG #3495: PostgreSQL does not use POSIX SHM or SEM in place of SYSV equivalents
Date
Msg-id 200707281709.l6SH9LEg087756@wwwmaster.postgresql.org
Whole thread Raw
Responses Re: BUG #3495: PostgreSQL does not use POSIX SHM or SEM in place of SYSV equivalents  (Heikki Linnakangas <heikki@enterprisedb.com>)
List pgsql-bugs
The following bug has been logged online:

Bug reference:      3495
Logged by:          Roger Leigh
Email address:      rleigh@whinlatter.ukfsn.org
PostgreSQL version: 8.2.4
Operating system:   Debian GNU/Linux (glibc 2.6, linux 2.6.22)
Description:        PostgreSQL does not use POSIX SHM or SEM in place of
SYSV equivalents
Details:

PostgreSQL is using the old SYSV shared memory and semaphores, as evidenced
by ipcs:

% sudo ipcs

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status
0x0052e2c1 0          postgres  600        29384704   2

------ Semaphore Arrays --------
key        semid      owner      perms      nsems
0x0052e2c1 0          postgres  600        17
0x0052e2c2 32769      postgres  600        17
0x0052e2c3 65538      postgres  600        17
0x0052e2c4 98307      postgres  600        17
0x0052e2c5 131076     postgres  600        17
0x0052e2c6 163845     postgres  600        17
0x0052e2c7 196614     postgres  600        17

------ Message Queues --------
key        msqid      owner      perms      used-bytes   messages


However, the SYSV memory limit is set in-kernel and while it can be easily
changed, it is rather less flexible than POSIX SHM support:

% df /dev/shm
Filesystem           1K-blocks      Used Available Use% Mounted on
tmpfs                   257584         0    257584   0% /dev/shm

On this system, that limit could be increased up to 6 GiB (though the above
would probably be the upper useful limit given that this is VM-backed).

Having the option of utilising either form of shared memory would be rather
useful.


Regards,
Roger

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: BUG #3494: may be Query Error: subplan does not executed
Next
From: Heikki Linnakangas
Date:
Subject: Re: BUG #3495: PostgreSQL does not use POSIX SHM or SEM in place of SYSV equivalents