I have pgpool talking to two PostgreSQL 8.3.14 servers.

One of the server gets all reads and writes (receives more CPU/IO load); the second server only gets writes (less load).

On the server which only receives writes, I noticed it has several "idle in transaction" processes, being several days old.

As I understand, "idle in transaction" are mostly due to ill-written software which doesn't close the transaction now and then.

The question is: why do I only see such long-living "idle in transaction" processes on one of the backends?

If at all, I should expect them on both backends, or?


--
Tomasz Chmielewski
http://wpkg.org
_______________________________________________
Pgpool-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgpool-general

Reply via email to