| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Donald Fraser" <demolish(at)cwgsy(dot)net> |
| Cc: | pgsql-admin(at)postgresql(dot)org |
| Subject: | Re: Lost connections |
| Date: | 2003-01-30 23:53:59 |
| Message-ID: | 706.1043970839@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-admin |
"Donald Fraser" <demolish(at)cwgsy(dot)net> writes:
> As you can see there still existed the connection to the database that I ha=
> d established from the Java app (from my computer 192.168.100.22).
FWIW, that backend will eventually time out and go away. The delay is
determined by whatever the TCP_KEEPALIVE code in your kernel does before
reporting the connection to be lost. It's probably an hour or so, if
the TCP code is RFC-compliant. I would not necessarily consider it a
good idea to shorten that delay; you'd be giving up robustness against
transient communication problems if you do.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Mintoo Lall | 2003-01-31 02:16:00 | Re: Postgres server output log |
| Previous Message | Tim Lynch | 2003-01-30 23:25:26 | block level db replication - eNBD? |