Slony 2.0.7 replication just stalled on big update on master (like
300.000 in batch.. nothing crazy).

Slony log on slave show repeatable (like each 5 minute) error:

2011-11-18 08:46:03 MSKERROR  remoteWorkerThread_4: "insert into
"_sports".sl_event     (ev_origin, ev_seqno, ev_timestamp,
ev_snapshot, ev_type     ) values ('4', '5000054127', '2011-11-18
06:18:44.863299', '922722:922722:', 'SYNC'); insert into
"_sports".sl_confirm     (con_origin, con_received, con_seqno,
con_timestamp)    values (4, 3, '5000054127', now()); insert into
"_sports".sl_event     (ev_origin, ev_seqno, ev_timestamp,
ev_snapshot, ev_type     ) values ('4', '5000054128', '2011-11-18
06:18:45.890297', '922731:922731:', 'SYNC'); insert into
"_sports".sl_confirm   (con_origin, con_received, con_seqno,
con_timestamp)    values (4, 3, '5000054128', now()); insert into
"_sports".sl_event     (ev_origin, ev_seqno, ev_timestamp,
ev_snapshot, ev_type     ) values ('4', '5000054129', '2011-11-18
06:18:46.921278', '922736:922736:', 'SYNC'); insert into
"_sports".sl_confirm      (con_origin, con_received, con_seqno,
con_timestamp)    values (4, 3, '5000054129', now()); commit
transaction;"
PGRES_FATAL_ERROR ERROR:  could not serialize access due to read/write
dependencies among transactions
DETAIL:  Reason code: Canceled on identification as a pivot, during write.
HINT:  The transaction might succeed if retried.
In that case slony just terminate without commiting batch (and
returned to the life only by watchdog).


Or (more rarely and seems harmless):
2011-11-18 08:49:42 MSKERROR  remoteWorkerThread_4: "update
"_sports".sl_setsync set     ssy_seqno = '5000054129', ssy_snapshot =
'922736:922736:',     ssy_action_list = '' where ssy_setid in
(2,3,5,4,7,6,8) and ssy_seqno < '5000054129'; "
ERROR:  could not serialize access due to read/write dependencies
among transactions
DETAIL:  Reason code: Canceled on identification as a pivot, during write.
HINT:  The transaction might succeed if retried.

Seem that new in 9.1 implementation of true serializable isolation
play here (and produce more serializability errors) and Slony not
ready to deal with serializability errors at that stage.

-- 
Maxim Boguk
Senior Postgresql DBA.

Phone RU: +7 910 405 4718
Phone AU: +61 45 218 5678

Skype: maxim.boguk
Jabber: maxim.boguk at gmail.com

LinkedIn profile: http://nz.linkedin.com/in/maximboguk
If they can send one man to the moon... why can't they send them all?

МойКруг: http://mboguk.moikrug.ru/
Сила солому ломит, но не все в нашей жизни - солома, да и сила далеко не все.