I have a long running finishTableAfterCopy() that is holding up slony's =

own delete. It has an AccessExclusiveLock on sl_config_lock, which the =

delete is trying to get an AccessExclusiveLock. This is with pg 8.4.1 =

and slonly 1.2.18. This is holding up my initial replication.


datid 	procpid 	usesysid 	usename 	current_query 	waiting 	xact_start =

query_start 	running time 	backend_start 	client_port
26836 	27144 	10 	postgres 	<IDLE> 	false 	=

	2009-12-08 18:14:34.302872 	=

	2009-12-08 17:18:34.093328 	52813
26836 	27159 	10 	postgres 	<IDLE> 	false 	=

	2009-12-08 18:15:02.770757 	=

	2009-12-08 17:19:58.528877 	60146
26836 	27165 	10 	postgres 	select "_pd".finishTableAfterCopy(29); =

analyze "public"."elog"; 	false 	2009-12-08 17:35:22.121016 	2009-12-08 =

18:06:49.681582 	505 secs 	2009-12-08 17:19:58.544084 	60147
26836 	27166 	10 	postgres 	<IDLE> 	false 	=

	2009-12-08 18:15:12.772765 	=

	2009-12-08 17:19:58.544504 	60149
26836 	27167 	10 	postgres 	delete from "_pd".sl_log_1 where log_origin =

=3D '1' and log_xid < '1531403324'; delete from "_pd".sl_log_2 where =

log_origin =3D '1' and log_xid < '1531403324'; delete from "_pd".sl_seqlog =

where seql_origin =3D '1' and seql_ev_seqno < '313199'; select =

"_pd".logswitch_finish(); 	true 	2009-12-08 17:35:22.124154 	2009-12-08 =

17:35:22.124154 	2,392.557 secs 	2009-12-08 17:19:58.545379 	60150
26836 	27168 	10 	postgres 	<IDLE> 	false 	=

	2009-12-08 18:14:18.765095 	=

	2009-12-08 17:19:58.546374 	60151
26836 	27716 	10 	postgres 	FETCH 1000 FROM c_postgres_status_bean =

false 	2009-12-08 18:15:14.664707 	2009-12-08 18:15:14.67136 	0.01 =

secs 	2009-12-08 18:15:14.660396 	37599

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.slony.info/pipermail/slony1-general/attachments/20091208/=
9f65386e/attachment.htm