On 2 November 2014 05:33, Mike Wilson wrote:
> Any recommendations would be very helpful.
Try using ionice and renice to increase the priority of the WAL sender
process on the master. If it helps, you are lagging because not enough
resources are being used by the sender process (rather than the
2014-11-02 19:16 GMT-02:00 Mike Wilson :
> Thanks for the information Greg.
>
> Unfortunately modifying the application stack this close to the holiday
> season won’t be an option so I’m left with:
>1) Trying to optimize the settings I have for the query mix I have.
>2) Optimize any long r
Thanks for the information Greg.
Unfortunately modifying the application stack this close to the holiday season
won’t be an option so I’m left with:
1) Trying to optimize the settings I have for the query mix I have.
2) Optimize any long running DML queries (if any) to prevent lag due to
l
Load on the slave is relatively light. It averages about 1.0 due to some data
ware house select queries running against it frequently. Previously only the
load on the master seems to have affected our replication lag no matter what
the slave was doing.
In thinking about this a bit more, the
Hello Mike,
what kind of load does the slave get?
what does the recovery process do on the slave during the times when lag is
being observed? Does it use 100% of the CPU?
WAL can be replayed by only one process, so no need to increase the
max_wal_senders.
Cheers,
-- Valentine Gogichashvili
On
I have two 9.3.4 PG instances that back a large internet website that has very
seasonal traffic and can generate large query loads. My instances are in a
master-slave streaming replication setup and are stable and in general perform
very well. The only issues we have with the boxes is that wh
