After a version upgrade (1.2.11 -> 1.2.15)
I now have all 3 servers at sony 1.2.15 (postgres version is 8.3 on
master1/slave2 and 8.2 on slave3).

slave3 ran out of disk space, but when I restarted slony it started
catching up. But now slave3 is in sync with node2, but both are behind
the master. Is there a way to get everything back to sync again?

log of master (after latest restart) looks like:

2009-01-21 15:56:26 CET CONFIG main: slon version 1.2.15 starting up
2009-01-21 15:56:26 CET DEBUG2 slon: watchdog process started
2009-01-21 15:56:26 CET DEBUG2 slon: watchdog ready - pid = 27145
2009-01-21 15:56:26 CET DEBUG2 slon: worker process created - pid = 27148
2009-01-21 15:56:27 CET CONFIG main: local node id = 1
2009-01-21 15:56:27 CET DEBUG2 main: main process started
2009-01-21 15:56:27 CET CONFIG main: launching sched_start_mainloop
2009-01-21 15:56:27 CET CONFIG main: loading current cluster configuration
2009-01-21 15:56:27 CET CONFIG storeNode: no_id=2 no_comment='Node 2 - '
2009-01-21 15:56:27 CET DEBUG2 setNodeLastEvent: no_id=2 event_seq=3705271
2009-01-21 15:56:27 CET CONFIG storeNode: no_id=3 no_comment='Node 3 - '
2009-01-21 15:56:27 CET DEBUG2 setNodeLastEvent: no_id=3 event_seq=2571462
2009-01-21 15:56:27 CET CONFIG storePath: pa_server=2 pa_client=1
2009-01-21 15:56:27 CET CONFIG storePath: pa_server=3 pa_client=1
2009-01-21 15:56:27 CET CONFIG storeListen: li_origin=2 li_receiver=1
li_provider=2
2009-01-21 15:56:27 CET CONFIG storeListen: li_origin=3 li_receiver=1
li_provider=2
2009-01-21 15:56:27 CET CONFIG storeListen: li_origin=3 li_receiver=1
li_provider=3
2009-01-21 15:56:27 CET CONFIG storeListen: li_origin=2 li_receiver=1
li_provider=3
2009-01-21 15:56:27 CET CONFIG storeSet: set_id=1 set_origin=1
set_comment='Set 1 for st1'
2009-01-21 15:56:27 CET DEBUG2 sched_wakeup_node(): no_id=1 (0 threads
+ worker signaled)
2009-01-21 15:56:27 CET CONFIG storeSet: set_id=2 set_origin=2
set_comment='Set 2 for st1'
2009-01-21 15:56:27 CET WARN   remoteWorker_wakeup: node 2 - no worker thread
2009-01-21 15:56:27 CET DEBUG2 sched_wakeup_node(): no_id=2 (0 threads
+ worker signaled)
2009-01-21 15:56:27 CET DEBUG2 main: last local event sequence = 3090600
2009-01-21 15:56:28 CET CONFIG main: configuration complete - starting threads
2009-01-21 15:56:28 CET DEBUG1 localListenThread: thread starts