Fri Sep 7 07:15:47 PDT 2007
- Previous message: [Slony1-general] "wait for event" invalidly requires node 1?
- Next message: [Slony1-general] failover doesn't remove triggers
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Dan, I was caught by this recently. There is a third parameter for "wait for event" which is the id of the node you are confirming has received the event. This defaults to 1 so you need to explicitly set it to the new origin node. You should also consider setting the 4th parameter which is a timeout. http://slony.info/documentation/stmtwaitevent.html Steve Hindmarch One IT - iBridge Development > -----Original Message----- > From: slony1-general-bounces at lists.slony.info > [mailto:slony1-general-bounces at lists.slony.info] On Behalf Of > Dan Falconer > Sent: 07 September 2007 14:47 > To: slony1-general at lists.slony.info > Subject: [Slony1-general] "wait for event" invalidly requires node 1? > > Running Postgres v8.0.13, Slony 1.2.11. In our > development replication > environment (thank God we finally got it setup), we've got > the following > setup (both running the same version of Slony + Postgres): > > Node 10 ("Cartman") > Debian 4.0 > > Node 20 ("Belial") > SLES 10.0 (x86_64) > > I had setup Belial to be the temporary master, then > attempted to move the > origin to Cartman. I used the "slonik_move_set" script, > provided with the > altperl tools, but decided to update it to contain the "wait > for event" > commands, as described on > http://slony.info/documentation/failover.html. > The following transcript of the session shows the problem with the > mysterious "node 1" error (with some added linebreaks to > accentuate each > run/output): > > --------------------------------- SNIP > --------------------------------- > postgres at belial:~/slony> vi bin/slonik_move_set > > postgres at belial:~/slony> ./bin/slonik_move_set --config > conf/slon_tools-pl.conf 1 20 10 > cluster name = dev_pl_replication; > node 10 admin conninfo='host=192.168.10.226 dbname=pl user=postgres > port=5432'; > node 20 admin conninfo='host=192.168.10.101 dbname=pl user=postgres > port=5432'; > echo 'Locking down set 1 on node 20'; > lock set (id = 1, origin = 20); > echo 'Locked down - moving it'; > wait for event (origin = 20, confirmed = 10); > move set (id = 1, old origin = 20, new origin = 10); > wait for event (origin = 20, confirmed = 10); > echo 'Replication set 1 moved from node 20 to 10. Remember to'; > echo 'update your configuration file, if necessary, to note the new > location'; > echo 'for the set.'; > > > postgres at belial:~/slony> ./bin/slonik_move_set --config > conf/slon_tools-pl.conf 1 20 10 | slonik > <stdin>:7: Error: No admin conninfo provided for node 1 > <stdin>:9: Error: No admin conninfo provided for node 1 > > > > postgres at belial:~/slony> ./bin/slonik_move_set --config > conf/slon_tools-pl.conf 1 20 10 | grep "node 1" > node 10 admin conninfo='host=192.168.10.226 dbname=pl user=postgres > port=5432'; > > > postgres at belial:~/slony> svn revert bin/slonik_move_set > Reverted 'bin/slonik_move_set' > > > postgres at belial:~/slony> ./bin/slonik_move_set --config > conf/slon_tools-pl.conf 1 20 10 > cluster name = dev_pl_replication; > node 10 admin conninfo='host=192.168.10.226 dbname=pl user=postgres > port=5432'; > node 20 admin conninfo='host=192.168.10.101 dbname=pl user=postgres > port=5432'; > echo 'Locking down set 1 on node 20'; > lock set (id = 1, origin = 20); > echo 'Locked down - moving it'; > move set (id = 1, old origin = 20, new origin = 10); > echo 'Replication set 1 moved from node 20 to 10. Remember to'; > echo 'update your configuration file, if necessary, to note the new > location'; > echo 'for the set.'; > > > postgres at belial:~/slony> ./bin/slonik_move_set --config > conf/slon_tools-pl.conf 1 20 10 | slonik > <stdin>:4: Locking down set 1 on node 20 > <stdin>:6: Locked down - moving it > <stdin>:8: Replication set 1 moved from node 20 to 10. Remember to > <stdin>:9: update your configuration file, if necessary, to > note the new > location > <stdin>:10: for the set. > -------------------------------- /SNIP > --------------------------------- > -- > Best Regards, > > > Dan Falconer > "Head Geek", Avsupport, Inc. / Partslogistics.com > http://www.partslogistics.com > _______________________________________________ > Slony1-general mailing list > Slony1-general at lists.slony.info > http://lists.slony.info/mailman/listinfo/slony1-general >
- Previous message: [Slony1-general] "wait for event" invalidly requires node 1?
- Next message: [Slony1-general] failover doesn't remove triggers
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list