On 11/15/19 17:38, Adrian Klaver wrote:
> On 11/15/19 1:54 PM, John Lumby wrote:
>> Adrian Klaver wrote :
>>>
>> We need to run with Read Committed.
>>
>> I am looking for a solution which does not alter the application or
>> overall behaviour,
>>
Adrian Klaver wrote :
>
> Seems you are looking for Serializable Isolation Level:
>
True , that would solve the race condition, but it is too drastic.
We need to run with Read Committed.
I am looking for a solution which does not alter the application or overall
behaviour,
but just addresses
Adrian Klaver wrote :
>
> On 11/15/19 10:37 AM, John Lumby wrote:
>
> > Suppose the original statement is
>
> > UPDATE myview VW set VW.counter = 11 where VW.primary_key = and
> > VW.counter = 10;
>
> > and my trigger constructs this s
I have an INSTEAD-OF row trigger one of whose purposes is to re-direct the
original statement
(which is operating on a VIEW) to a different real base table.
Suppose the original statement is
UPDATE myview VW set VW.counter = 11 where VW.primary_key = and
AND VW.counter = 10;
an
> From: Peter Geoghegan
> Sent: July 9, 2019 5:15 PM
> Subject: Re: REINDEX : new parameter to preserve current average leaf density
> as new implicit FILLFACTOR
>
> On Tue, Jul 9, 2019 at 12:29 PM John Lumby wrote:
> > I was not thinking of a new command, just an e
> From: Peter Geoghegan
> Sent: July 9, 2019 3:01 PM
> Subject: Re: REINDEX : new parameter to preserve current average leaf density
> as new implicit FILLFACTOR
>
> On Tue, Jul 9, 2019 at 11:27 AM John Lumby wrote:
> > And the point of the REINDEX at that point (below)
t average leaf density
> as new implicit FILLFACTOR
>
> On Tue, Jul 9, 2019 at 10:31 AM John Lumby wrote:
> > Yes, I see that. But surely "making splits occur less often" is a
> > desirable
> > objective in itself, is it not? And I believe that
> From: Peter Geoghegan
> Sent: July 9, 2019 1:47 PM
> Subject: Re: REINDEX : new parameter to preserve current average leaf density
> as new implicit FILLFACTOR
>
> On Tue, Jul 9, 2019 at 10:31 AM John Lumby wrote:
> > Yes, I see that. But surely "making
> From: Peter Geoghegan
> Sent: July 8, 2019 1:39 PM
> Subject: Re: REINDEX : new parameter to preserve current average leaf density
> as new implicit FILLFACTOR
>
> On Mon, Jul 8, 2019 at 9:23 AM John Lumby wrote:
>
> > Although a welcome improvement, I think
| 0 | ms
autovacuum_vacuum_cost_limit| |
autovacuum_vacuum_scale_factor | 0 |
autovacuum_vacuum_threshold | 2000|
autovacuum_work_mem | 1048576 | kB
To correspond to your " more churn between each VACUUM"
Would you then suggest increasing
autovacuum_vacuum_
e
previous statistics collection that can be inserted without incurring an index
scan at the VACUUM cleanup stage.
I have not researched this at all and nor did I set it to anything for my
pg-12beta2 run, but it sounds as though maybe it could be relevant to this
kind of workload - Is that so?
CheersJohn Lumby
p5RaATILoiE
>
Is there a pdf or text version?
>
> Does it seem at all possible that you were affected by either the issue with
> duplicates,
>
definitely
>
> or the issue that is addressed by the "split after new tuple" optimization?
>
don't know,
nge much if the workload is truly steady-state), but I wonder
if this community would agree that it would provide a useful automation of the
process.
Cheers, John Lumby
13 matches
Mail list logo
