Hi Christian:
On Wed, 29 May 2024 at 00:59, Christian Schröder
<christian.schroeder@wsd.com> wrote:
> Thank you for your advice. I used "ipcs" to get more readable information about the shared memory:
...
> As far as I understand, there is no upper limit to the size of the shared memory. The database only holds a single
sharedmemory segment, which doesn't seem to have a relevant size.
Seems the same to me, so I will disregard that.
> I am surprised to see this since I would have expected much more shared memory to be used by the database. Is there
anythingin the configuration that prevents the shared memory from being used?
I am not too current with postgres, that one was a thing which
happened to me when I did more administration, and is one you always
want to check. I assume you have all checked, but I would follow by
insuring every mounted partition in your system has space. I am not
current on the details, but I know Pg can be mmaping things and doing
other stuff. Your problem seems more of resource exhaustion, so I
would follow by checking that, both disk, tmpfs and all the other
stuff. I cannot give you advice on that as it depends a lot on your
server configuration and from the age in the mssages I suspect you
have the usual suspects debugged. But as you have a configuration
crahsing in minutes and it seems to be a dev server you could do it
easily.
Sorry.
Francisco Olarte.