On Sat, 13 Nov 2010, Aleksey Tsalolikhin wrote:

> Replication lag is present (and growing) again after resync.  Help!
>
> We do use check_postgres.pl, thanks, Brian, that's how we noticed the problem.
>
> I've reviewed http://www.slony.info/documentation/1.2/loganalysis.html,
> thanks, Steve; there are no ERROR messages on the slon slave or
> master, only DEBUG.
>
> There is one interesting DEBUG message on the master, and I could not
> find it in loganalysis.html:  "calc sync size", and the ideal is often
> larger than the proposed, example:  (ideal: 14, proposed: 3)
>
>
> DEBUG2 remoteListenThread_2: queue event 2,2659 SYNC
> DEBUG2 remoteWorkerThread_2: Received event 2,2659 SYNC
> DEBUG2 calc sync size - last time: 1 last length: 4152 ideal: 14
> proposed size: 3
> DEBUG2 remoteWorkerThread_2: SYNC 2659 processing
> DEBUG2 remoteWorkerThread_2: no sets need syncing for this event
>
>
> another example:  (ideal 11, proposed 5)
>
>
> DEBUG2 remoteListenThread_2: UNLISTEN
> DEBUG2 remoteWorkerThread_2: Received event 2,2667 SYNC

This line means that the 2667'th event originating from node 2 has been 
processed (I am assuming that this is the node 1 log).

In the node to log do you see similar lines but with an event like "1,1234"
indicating that node 2 is processing events from node 1.

If not what does show up in your log for that node.

> DEBUG2 calc sync size - last time: 2 last length: 10008 ideal: 11
> proposed size: 5



> DEBUG2 remoteWorkerThread_2: SYNC 2667 processing
> DEBUG2 remoteWorkerThread_2: no sets need syncing for this event
> DEBUG2 localListenThread: Received event 1,14742 SYNC
>
>
> Would suboptimal sync size influence replication time?
>
> What else can I do to analyze and resolve this problem, please?
> _______________________________________________
> Slony1-general mailing list
> Slony1-general at lists.slony.info
> http://lists.slony.info/mailman/listinfo/slony1-general
>