Tom, thanks very much for your response.
I've run REINDEX on the db and ANALYZE multiple times - there is no
dramatic improvement.
I appreciate all the impressive effort and contribution of PostgreSQL
developers and am fully aware of importance of bug fixes. Please
acknowledge I am running into
On 2012-05-18, Jon Smark wrote:
> I can think of two solutions:
>
> 1) "SELECT wid, data FROM widgets WHERE wid IN $targets"
> 2) "SELECT wid, data FROM widgets WHERE ARRAY [wid] <@ $targets"
>
my testing indicates that from 1 is significantly faster than from 2
postgres (8.4) cant factor "ARRAY
Finally... issue is resolved :) Increased the DST to 10K and ran ANALYZE
on the db again.
QUERY PLAN
--
Sort (cost=7561.71..7562
On 05/18/2012 04:19 PM, David Salisbury wrote:
I'm trying to debug an intermittent problem I'm seeing in one of our
rollup scripts.
I'll try to summarize. A table has a measured_at field, of which I
calculate another
time value based on that field and a longitude value, called solar_noon,
and I
Hi again,
We are still constantly getting postgresql processes killed by signal
9 from time to time, without any idea why or how.
Syslog seems completly clean.
In case a postgresql process would exceed some restricted resources
like file descriptors, would the kernel choose to terminate it using
Hi,
In day to day work I need to dump objects of only specific types like
functions, types, view etc from a schema. I know that it will be better to
use separate schemas for similar tasks and already using it. This problem
arises when I have to issue an update script to the client having only
func
On Sat, May 19, 2012 at 3:09 AM, Steve Crawford
wrote:
> I also don't understand the "xcrm.channel_id in (1)" instead of
> "xcrm.channel_id = 1" unless this is a generated query and there could be
> multiple ids in that condition.
Side point from this thread. Is there ever any difference between
On 05/19/2012 10:34 AM, David Salisbury wrote:
CCing the list.
On 5/19/12 8:12 AM, Adrian Klaver wrote:
I hope no one looks further into the problem as the case is closed. It
was a coding
problem and not a time matchup problem. Late Friday afternoons just
aren't my most
shining moments. ;
On Fri, May 18, 2012 at 5:58 PM, Bosco Rama wrote:
> The per-blob messages are missing whether we use the '-j2' or not and also
> regardless of number of '-v's we use.
>
> I'm glad they've reinstated some of the messages. Previously there was
> a message per blob between the "restoring BLOBS" an
Hey Josh,
I found the message I was seeing. It was/is(?) in StartRestoreBlob() and it
looks like this:
ahlog(AH, 2, "restoring large object with OID %u\n", oid);
But I don't know how to find it in the current git tree or how to activate
it from the command-line (assuming it is still part of
10 matches
Mail list logo
