JP Fletcher jpfletch
Wed Oct 26 20:22:39 PDT 2005
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I'm testing v 1.1.2 on 7.4.8.  In a simple 3 node cluster with a single
set, if I move the set from provider to one of the two subscribers, the
slon for the other subscriber shuts down,  for instance:


1.  node1 is provider to subscribers node2 and node3.

2.  slonik 'move set' moves set from node1 to node2.

3.  following receipt of 'move set', 'accept set', slon for node3 shuts down

Here's what the log has to say:

2005-10-26 18:55:40 UTC DEBUG2 remoteWorkerThread_1: Received event
1,10217 MOVE_SET
2005-10-26 18:55:40 UTC DEBUG2 remoteWorkerThread_2: forward confirm
1,10216 received by 2
2005-10-26 18:55:40 UTC CONFIG moveSet: set_id=1 old_origin=1 new_origin=2
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=1 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=2 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC DEBUG1 remoteWorkerThread_2: helper thread for
provider 1 created
2005-10-26 18:55:40 UTC CONFIG storeListen: li_origin=1 li_receiver=3
li_provider=1
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=1 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC CONFIG storeListen: li_origin=2 li_receiver=3
li_provider=1
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=1 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=1 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC DEBUG2 sched_wakeup_node(): no_id=2 (0 threads +
worker signaled)
2005-10-26 18:55:40 UTC DEBUG1 remoteWorkerThread_1: helper thread for
provider 1 terminated
2005-10-26 18:55:40 UTC DEBUG1 remoteWorkerThread_1: disconnecting from
data provider 1
2005-10-26 18:55:40 UTC DEBUG2 remoteListenThread_1: start listening for
event origin 2
2005-10-26 18:55:40 UTC INFO   remoteListenThread_2: disconnecting from
'host=localhost dbname=oxrsorg_node2 user=oxrsorg port=5636
password=oxrsorg'
2005-10-26 18:55:40 UTC DEBUG1 remoteListenThread_2: thread done
2005-10-26 18:55:43 UTC DEBUG2 remoteListenThread_1: queue event 2,7960
ACCEPT_SET
2005-10-26 18:55:43 UTC DEBUG2 remoteWorkerThread_2: Received event
2,7960 ACCEPT_SET
2005-10-26 18:55:43 UTC DEBUG2 start processing ACCEPT_SET
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT: set=1
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT: old origin=1
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT: new origin=2
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT: move set seq=7960
2005-10-26 18:55:43 UTC DEBUG2 got parms ACCEPT_SET
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT_SET - node not origin
2005-10-26 18:55:43 UTC DEBUG2 remoteWorkerThread_1: forward confirm
1,10217 received by 2
2005-10-26 18:55:43 UTC DEBUG2 ACCEPT_SET - MOVE_SET or FAILOVER_SET
exists - done
2005-10-26 18:55:43 UTC DEBUG2 remoteWorkerThread_1: forward confirm
2,7960 received by 1
2005-10-26 18:55:43 UTC DEBUG1 slon: shutdown requested
2005-10-26 18:55:43 UTC DEBUG2 slon: notify worker process to shutdown
2005-10-26 18:55:43 UTC DEBUG2 slon: wait for worker process to shutdown
2005-10-26 18:55:43 UTC INFO   remoteListenThread_1: disconnecting from
'host=localhost dbname=oxrsorg_node1 user=oxrsorg port=5634
password=oxrsorg'
2005-10-26 18:55:43 UTC DEBUG1 main: scheduler mainloop returned
2005-10-26 18:55:43 UTC DEBUG2 main: wait for remote threads
2005-10-26 18:55:43 UTC DEBUG1 syncThread: thread done
2005-10-26 18:55:43 UTC DEBUG1 remoteListenThread_1: thread done
2005-10-26 18:55:43 UTC DEBUG2 sched_wakeup_node(): no_id=1 (0 threads +
worker signaled)
2005-10-26 18:55:43 UTC DEBUG1 cleanupThread: thread done
2005-10-26 18:55:43 UTC DEBUG1 localListenThread: thread done
2005-10-26 18:55:43 UTC DEBUG1 remoteWorkerThread_1: thread done
2005-10-26 18:55:43 UTC DEBUG2 sched_wakeup_node(): no_id=2 (0 threads +
worker signaled)
2005-10-26 18:55:43 UTC DEBUG2 main: notify parent that worker is done
2005-10-26 18:55:43 UTC DEBUG1 main: done
2005-10-26 18:55:43 UTC DEBUG2 slon: worker process shutdown ok
2005-10-26 18:55:43 UTC DEBUG2 slon: exit(-1)


This doesn't seem to cause much of a problem, as the watchdog process
starts the slon back up.  Is this the expected behaviour?

- --
JP Fletcher
Database Administrator
Afilias Canada
voice: 416.646.3304 ext. 123
fax: 416.646.3305
mobile: 416.561.4763
jpfletch at ca.afilias.info
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFDX9j+aQwDW0M+H30RAtdTAJ4+8F25pHjaTpAqu0a5/oe+it3oOACdEUcs
ww+sknxTAJdNIE2irLNlL70=
=Me9U
-----END PGP SIGNATURE-----


More information about the Slony1-general mailing list