Hi folks,

 

I have a 3-node system, node1 replicates to node2 replicates to node3,
each node running Slony 1.1.5

 

Today I tried adding a table to replication.

 

-          I created the table on nodes 1, 2, and 3

-          I created a new set including the new table (set2) using
slonik_create_set set2

-          I then ran slonik_subscribe_set set2 node2

-          I then ran slonik_subscribe_set set2 node3

 

The subscribe set worked just fine on node2, but not on node3. I'm
thinking the problem may stem from the perl script setting the provider
to be node 1 instead of node 2 when I ran subscribe set for node 3.

 

At any rate here's what's showing up in the Slony logs:

 

2006-10-17 17:42:32 MDT ERROR  remoteWorkerThread_1: "declare LOG cursor
for select     log_origin, log_xid, log_tableid,     log_actionseq,
log_cmdtype, log_cmddata from "_namesdb_production".sl_log_1 where
log_origin = 1 and (  order by log_actionseq; " PGRES_FATAL_ERROR ERROR:
syntax error at or near "order" at character 177

 

2006-10-17 17:42:32 MDT ERROR  remoteWorkerThread_1: "close LOG; "
PGRES_FATAL_ERROR ERROR:  current transaction is aborted, commands
ignored until end of transaction block

 

I'd like to just drop the replication for set2 and start over again if
that's possible, but I'm guessing Slony on node3 isn't ever going to see
the drop set command since it's hung up on the above.

 

Any thoughts or suggestions on how best to proceed here? I'd really
rather not break the entire replication setup and start from scratch. If
someone could point me to the right data I can munge in the PG schema
that would correct the above errors if possible, I'd be grateful.

 

Thanks for your help,

 

-- Dan Hanks

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gborg.postgresql.org/pipermail/slony1-general/attachments/20061017/0d2f020a/attachment.html