Re: postgres memory management

Hi Clift,

you are right, I have the same problem on RedHat. After I inserted -e it
works so far. But there's something else that seems strange to me I'm not
quite sure if I'm reading this right since I understand only half of what
happens in this script. After the comment that says "Don't do anything if
process still running..." on line there is the following sequence of lines:

ps hj$ipcs_pid >/dev/null 2>&1
if [ $? -eq 0 ]; then
echo "skipped...."

As I understand it the if statement tests the output of the previous ps
statement. The strange thing is that the variable $ipcs_pid is never used
anywhere before this line, so I think it's always null (or whatever this
scripting language defaults to). There are three other variables ipcs_id,
ipcs_cpid and ipcs_lpid but no ipcs_pid. If I'm right here, it seems that
this script does effectively nothing in terms of shared memory.

Please tell me if I'm on a completely wrong track :-)

Alexander Jerusalem
ajeru(at)gmx(dot)net
vknn

At 03:03 23.01.01, Justin Clift wrote:
>Hi Alexander,
>
>I've noticed that the PG 7.03 ipcclean script uses "ps x | grep -s
>'postmaster'" to determine if a postmaster daemon is still running,
>which at least for Mandrake Linux 7.2 doesn't work as expected. With
>this version of linux, the ps & grep combination will find itself and
>then ipcclean will complain about an existing postmaster.
>
>I found the solution to this being to edit the ipcclean script and
>change the "ps x | grep -s 'postmaster'" part to "ps -e | grep -s
>'postmaster'". This then works correctly with Mandrake 7.2.
>
>Regards and best wishes,
>
>Justin Clift
>
><snip>
> >
> > Oddly, when I try to run ipcclean a second time, it says: ipcclean: You
> > still have a postmaster running. Which is not the case as ps -e proves.
> >
> > Alexander Jerusalem
> > ajeru(at)gmx(dot)net
> > vknn