are in semop
semop(2064417, {{0, -1, 0}}, 1^C
ipcs reveals no shortage of semaphores..
Please any ideas very welcome, our service now in a very sad state..
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>M: +30 6973-903808T: +30 210-6802-358*
Os is debian 6
help debugging.
Is there a way to downgrade postgresql ?
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>M: +30 6973-903808T: +30 210-6802-358*
On 20 July 2015 at 01:01, Tom Lane wrote:
> Spiros Ioannou writes:
> > We upgraded to 9.4.4 from 8.
mp/buildd/postgresql-9.4-9.4.4/build/../src/backend/postmaster/postmaster.c:1244
#15 0x7f47b24ba474 in main (argc=5, argv=0x7f47b49b4920) at
/tmp/buildd/postgresql-9.4-9.4.4/build/../src/backend/main/main.c:228
Detaching from program: /usr/lib/postgresql/9.4/bin/postgres, process 15601
Debian 6, x64, HP Proliant, dbsize: 3TB on EMC AX-4i (iscsi), 16 cores,
24GB RAM.
kernel: Linux vserver11 2.6.32-5-xen-amd64 #1 SMP Sun Dec 7 22:52:42 UTC
2014 x86_64 GNU/Linux
I attach postgresql.conf
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>
FYI we have an 9.3.5 with commit_delay = 4000 and commit_siblings = 5 with
a 8TB dataset which seems fine. (Runs on different - faster hardware
though).
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>M: +30 6973-903808T: +30 210-6802-358*
On 20 July 2
has about 5GB)
* WITH d_rows AS (DELETE FROM measurement_events_a RETURNING * ) INSERT
INTO measurement_events_b SELECT * FROM d_rows;
The above caused the problem to appear every time, after a 10-20 minutes.
Regards,
-Spiros
*Spiros Ioannou IT Manager, inAccesswww.inaccess
>
> Hmm. With that CTE query, were there other queries running at the same
> time?
yes, a multitude of INSERT and one COMMIT as I remember. I've noticed that
in every stuck situation there was always 1 stuck COMMIT running. We do
mainly INSERT and SELECT, very rarely UPDATE or DELETE.
but the stu
That would need a replica of the data probably which is not possible
(tablespace is 4TB).
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>M: +30 6973-903808T: +30 210-6802-358*
On 30 July 2015 at 21:47, Scott Marlowe wrote:
> You might want to try
srcs_timestamps mt (cost=0.00..5733.57
rows=350257 width=24)
(6 rows)
We have tried so far fiddling with work_mem up to 512M - no difference.
Any suggestions?
Thanks for any help,
-Spiros Ioannou
inaccess
on wrote:
> On 7/8/2014 4:47 AM, Spiros Ioannou wrote:
>
>> While executing the following query through psql :
>>
>> SELECT me.* FROM measurement_events me JOIN msrcs_timestamps mt ON
>> me.measurement_source_id=mt.measurement_source_id WHERE
>> measureme
Hello group,
we have a timeseries table, and we tried to partition it by month (with
pg_partman). It seems the query planner always reads all tables regardless
of WHERE, except when WHERE is equality.
*the parent table:*
ifms_db=# \dS measurement_events
Table "public.measurement_eve
@Rob Sargent: sorry Rob, not sure what you are asking.
@Kyotaro HORIGUCHI
thanks for your reply and time Kyotaro,
Using the following query
EXPLAIN ANALYZE SELECT * FROM measurement_events WHERE
measurement_source_id='df86917e-8df0-11e1-8f8f-525400e76ceb' AND
measurement_time >= '2015-01-01 00:0
>
>
> > EXPLAIN ANALYZE SELECT * FROM measurement_events WHERE
> > measurement_source_id='df86917e-8df0-11e1-8f8f-525400e76ceb' ORDER BY
> > measurement_time DESC LIMIT 1;
> >
> > This seems to fail, scanning all tables. Do you think this can be
> improved
> > at all ? The query plan of the above
-Spiros
*Spiros Ioannou IT Manager, inAccesswww.inaccess.com
<http://www.inaccess.com>M: +30 6973-903808T: +30 210-6802-358*
On 22 January 2015 at 17:37, Tom Lane wrote:
> Spiros Ioannou writes:
> > It is ORDER BY measurement_time, not measurement_id, and measurement_time
I'm writing a c-language user defined function for postgresql 9.3.5. I
didn't use pl/pgsql or pl/python because part of it needs to be written in
C for speed.
The function gets an existing OID as an argument. How can I access & modify
the OID data from within the c-language function? I can't find
But these need a connection to the postgresql backend. It seems an overkill
for postgresql to connect to itself, I would prefer something more optimal.
It seems for now that reading libpq's source code is the only way to go.
Regards,
Spiros
Thanks, that's interesting, I will look into that.
Regards,
-Spiros
We are dumping a 4TB postgres database using pg_dump. The source is on a
copy-on-write snapshot filesystem, and the target is an external drive.
While the pg_dump is running we get about 4-5MB/sec writes on the
filesystem where postgresql data is (iostat), and specifically inside the
base/ directo
gt; On Mon, Dec 2, 2013 at 6:22 AM, Spiros Ioannou wrote:
>
>> We are dumping a 4TB postgres database using pg_dump. The source is on a
>> copy-on-write snapshot filesystem, and the target is an external drive.
>>
>> While the pg_dump is running we get about 4-5MB/sec write
r the info.
On 10 December 2013 21:13, Jeff Janes wrote:
> On Tue, Dec 10, 2013 at 12:43 AM, Spiros Ioannou wrote:
>
>> Hi Jeff,
>> autovacuum is off on the DB running on the filesystem snapshot.
>>
>
>
> Which probably makes sense on the snapshot, but is it a
ut 130MB large and the mysql's one
was 64MB, is this normal?
Thank you for your time,
Spiros Ioannou
e-mail:[EMAIL PROTECTED]
---
Image Video & Multimedia Systems Lab.
Department of Electrical &
22 matches
Mail list logo