Ah... no I didn't know that - that would explain all the other
behaviour then!! Good to know.
Alex
On Sat, Mar 8, 2008 at 3:03 PM, Tom Lane <[EMAIL PROTECTED]> wrote:
> "Alex Turner" <[EMAIL PROTECTED]> writes:
> > ... I found what was going on in that bit and fixed it so
>
> > it's not crashin
"Alex Turner" <[EMAIL PROTECTED]> writes:
> ... I found what was going on in that bit and fixed it so
> it's not crashing anymore, I'm just worried about how on earth that
> could have affected other back end processes that were querying
> unrelated databases.
You do know that a crash in any backe
On Sat, Mar 8, 2008 at 1:05 PM, Douglas McNaught <[EMAIL PROTECTED]> wrote:
> On 3/8/08, Alex Turner <[EMAIL PROTECTED]> wrote:
>
> > No I'm not. Where would a core file be if there was going to be one?
>
> They should appear in the data directory (e.g. /var/lib/pgsql/data).
>
Yeah - thats where
On 3/8/08, Alex Turner <[EMAIL PROTECTED]> wrote:
> No I'm not. Where would a core file be if there was going to be one?
They should appear in the data directory (e.g. /var/lib/pgsql/data).
> I'm not sure how I can tell if the ulimit applies to the running
> postmaster
>
> I am the postgres u
No I'm not. Where would a core file be if there was going to be one?
I'm not sure how I can tell if the ulimit applies to the running
postmaster
I am the postgres user and ulimit -a show unlimited for core, and I
run pg_ctl start. I have put it in that one place in /etc/ and also
in ~/.bash_prof
On 3/8/08, Alex Turner <[EMAIL PROTECTED]> wrote:
> Well - I know that my stored proc is segfaulting based on a strace of
> postgresql. Don't know how that affects trac which isn't using that
> stored proc... the mystery continues. Either way I didn't get a
> corefile, and ulimit -a show I hav
Well - I know that my stored proc is segfaulting based on a strace of
postgresql. Don't know how that affects trac which isn't using that
stored proc... the mystery continues. Either way I didn't get a
corefile, and ulimit -a show I have unlimited core file size :(
Alex
On Fri, Mar 7, 2008 at 1
Well - I think it might be that some of my servlets weren't closing
their database connections properly.
I do have some new evidence though:
I did an strace of the tomcat processes, and I noticed something that
might be odd, but I'm not really qualified to say. I notice that
every time a socket
On Fri, Mar 7, 2008 at 11:17 AM, Alex Turner <[EMAIL PROTECTED]> wrote:
> I didn't. And after the reboot, I still see 8 new sockets stuck in
> CLOSE_WAIT - I'm wondering if this is a hardware/kernel problem...
Having sockets in CLOSE_WAIT is actually pretty normal
--
Sent via pgsql-general mai
I didn't. And after the reboot, I still see 8 new sockets stuck in
CLOSE_WAIT - I'm wondering if this is a hardware/kernel problem...
Alex
On Fri, Mar 7, 2008 at 11:50 AM, Alex Turner <[EMAIL PROTECTED]> wrote:
> Nothin worth mentioning in /var/log/messages
>
> The wierd thing I do see is there
Nothin worth mentioning in /var/log/messages
The wierd thing I do see is there are a number of sockets in
CLOSE_WAIT when doing a netstat -an | grep 5432
I think maybe I'll just reboot and see if that fixes it.
Alex
On Fri, Mar 7, 2008 at 11:47 AM, Richard Huxton <[EMAIL PROTECTED]> wrote:
> Al
Alex Turner wrote:
It was core dumping on the 5th of March, but it hasn't since. It's
just failing with the connection closed problem. It seems to happen
worst with queries that are going to do updates and with connections
that are persistent between http requests...
I downgraded to 8.2, but i
It was core dumping on the 5th of March, but it hasn't since. It's
just failing with the connection closed problem. It seems to happen
worst with queries that are going to do updates and with connections
that are persistent between http requests...
I downgraded to 8.2, but it hasn't made any dif
Alex Turner wrote:
I had stored procs in C on 8.2 for months, and I moved them over to
8.3 when we upgraded.
And re-compiled them, yes?
> The thing is that it's happening on a database
that doesn't have the stored procs. It seems to be happening worst on
a system that keeps connections open
I did a pg_ctl start from the postgres user... is that gonna work, or
does pg_ctl do an su?
Alex
On Thu, Mar 6, 2008 at 4:34 PM, Tom Lane <[EMAIL PROTECTED]> wrote:
> "Alex Turner" <[EMAIL PROTECTED]> writes:
> > I'm thinking I need a core file.
>
> Yeah.
>
>
> > I'm guessing I just restart
>
"Alex Turner" <[EMAIL PROTECTED]> writes:
> I'm thinking I need a core file.
Yeah.
> I'm guessing I just restart
> postgresql from a user whos core file size limit is set to non zero?
Probably won't help if your start script does "su postgres" as most do.
What I'd try is putting the ulimit comma
Alex Turner wrote:
Ok - lookint at the pg log, it appears that the server process is seg
faulting :(. This might conceivably be my fault. I have 3 stored
procedures written in C, but they've been on the server for months,
and unless I didn't deploy them correctly originally I don't think
it's t
On 3/6/08, Alex Turner <[EMAIL PROTECTED]> wrote:
> Ok - lookint at the pg log, it appears that the server process is seg
> faulting :(. This might conceivably be my fault. I have 3 stored
> procedures written in C, but they've been on the server for months,
> and unless I didn't deploy them c
Ok - more info
I turned on connection logging a few other things, and normaly I get
something like this:
LOG: 0: connection authorized: user=postgres database=trend
LOCATION: BackendInitialize, postmaster.c:3097
LOG: 0: disconnection: session time: 0:00:00.004 user=postgres
database=tr
Sometimes I'm getting LOG: unexptected EOF on client connection
Alex
On Thu, Mar 6, 2008 at 4:06 PM, Alex Turner <[EMAIL PROTECTED]> wrote:
> Ok - the connection closed thing is happening a lot, but not much is
> going into pg_log...
>
> Alex
>
>
>
> On Thu, Mar 6, 2008 at 4:03 PM, Douglas McN
Alex Turner wrote:
I'm getting the back end closing connections early for some reason.
Here is an exception report from my servlet. This first started
happening with my instance of Trac, but now it's happening to my Java
apps too. I hope someone can shed some light on what is going on
here.
W
Ok - the connection closed thing is happening a lot, but not much is
going into pg_log...
Alex
On Thu, Mar 6, 2008 at 4:03 PM, Douglas McNaught <[EMAIL PROTECTED]> wrote:
> On 3/6/08, Alex Turner <[EMAIL PROTECTED]> wrote:
> > Ok - lookint at the pg log, it appears that the server process is seg
On 3/6/08, Alex Turner <[EMAIL PROTECTED]> wrote:
> I'm getting the back end closing connections early for some reason.
> Here is an exception report from my servlet. This first started
> happening with my instance of Trac, but now it's happening to my Java
> apps too. I hope someone can shed
Ok - lookint at the pg log, it appears that the server process is seg
faulting :(. This might conceivably be my fault. I have 3 stored
procedures written in C, but they've been on the server for months,
and unless I didn't deploy them correctly originally I don't think
it's them.
I'm thinking I
I'm getting the back end closing connections early for some reason.
Here is an exception report from my servlet. This first started
happening with my instance of Trac, but now it's happening to my Java
apps too. I hope someone can shed some light on what is going on
here.
Alex
HTTP Status 500
25 matches
Mail list logo
