initdb error message

Am Freitag, 23. Mï¿œrz 2007 11:44 schrieb Susanne Ebrecht:

I tried an initdb and got a error message about no space left on device
and semaphores stuff (the message was in german, because my system is
german, that's the reason, I don't paste this message here).

No hint, that I just have forgotten, to stop the daemon.

Is it possible to add a hint in that error message, that maybe just
the daemon is still running?

If a daemon is still running in the same directory, you will get an error
message that the directory still exists. Therefore, the daemon must be
running in a different directory, and there is no definite reason why you
should stop that one. It might be serving some other application.

--
Peter Eisentraut
http://developer.postgresql.org/~petere/

Peter Eisentraut wrote:

Am Freitag, 23. Mï¿œrz 2007 11:44 schrieb Susanne Ebrecht:

I tried an initdb and got a error message about no space left on device
and semaphores stuff (the message was in german, because my system is
german, that's the reason, I don't paste this message here).

No hint, that I just have forgotten, to stop the daemon.

Is it possible to add a hint in that error message, that maybe just
the daemon is still running?

If a daemon is still running in the same directory, you will get an error
message that the directory still exists. Therefore, the daemon must be
running in a different directory, and there is no definite reason why you
should stop that one. It might be serving some other application.

Thanks Peter :)
That was, what I first thought, that the daemon must not be stopped, if
I want to do an initdb to another directory. But the guys on irc called
me weird. They said, my thought were wrong.

Today, in the afternoon, I couldn't run two clusters at the same time.
Therefor we figured the problem out and I found the same error message
at /var/log/messages.

Just setting max_connections to a lower value solved the problem.

So, it's no bug, just a kind of making me weird. :)

Susanne