Re: BUG #8404: JDBC block hot standby replication? - Mailing list pgsql-bugs

From Heikki Linnakangas
Subject Re: BUG #8404: JDBC block hot standby replication?
Date
Msg-id 521DEDD2.7010405@vmware.com
Whole thread Raw
In response to Re: BUG #8404: JDBC block hot standby replication?  (Valentine Gogichashvili <valgog@gmail.com>)
List pgsql-bugs
On 28.08.2013 12:46, Valentine Gogichashvili wrote:
> Hello
>
> This is a well documented feature of Hot-Standby Replication.
>
> see:
> http://www.postgresql.org/docs/9.2/static/runtime-config-replication.html#RUNTIME-CONFIG-REPLICATION-STANDBY

That explains how streaming delay and query cancellations work, but the
OP's question was why running the same statements over JDBC behaves
differently than running them over psql.

I'm afraid I don't know the answer. One guess is that when you make the
JDBC connection, you have asked for repeatable read or serializable
isolation level, while psql runs in read committed mode. Running a "show
transaction_isolation" in both would show if that's the case. I'd also
suggest doing "select * from pg_stat_activity" to see if the session
looks the same in both cases.

- Heikki

pgsql-bugs by date:

Previous
From: Valentine Gogichashvili
Date:
Subject: Re: BUG #7494: WAL replay speed depends heavily on the shared_buffers size
Next
From: Heikki Linnakangas
Date:
Subject: Re: BUG #7494: WAL replay speed depends heavily on the shared_buffers size