On 24/03/17 11:23, Erik Rijkers wrote:
> On 2017-03-24 10:45, Mark Kirkwood wrote:
>>
>> However one minor observation - as Michael Banck noted - the elapsed
>> time for slave to catch up after running:
>>
>> $ pgbench -c8 -T600 bench
>>
>> on the master was (subjectively) much longer than for physical
>> streaming replication. Is this expected?
>>
>
> I think you probably want to do (on the slave) :
>
> alter role <you role> set synchronous_commit = off;
>
> otherwise it's indeed extremely slow.
>
Yes that might be one reason, see
https://www.postgresql.org/message-id/08b7053b-5679-0733-3af7-00b8cde62c90@2ndquadrant.com
(although it probably needs rebase now)
Also don't forget that the snapbuild performance fixes haven't been
committed yet so it can take long time to even start.
-- Petr Jelinek http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training &
Services