Would like to try out the newest version for testing
The init.d/pgcluster script seems to be missing start_recover from the opts.
After adding it, I tried running start_recover, first it asks if I want to add
the key, I responded with yes, then it says:
* Starting PGCluster...
Response: * Please see log file: /var/lib/postgresql/data/postgresql.log
And here is what is in that log:
Start in recovery mode!
Please wait until a data synchronization finishes from Master DB...
1st recovery step of [global] directory...NG
PGR_Get_Cluster_Conf_Data failedPermission denied.
rsync: connection unexpectedly closed (0 bytes read so far)
rsync error: error in rsync protocol data stream (code 12) at io.c(189)
I setup the ssh keys as instructed on the pgcluster site - although I had setup
sshd to allow protocol 1 and change it back to the default port 22.
I'm not sure what to check next. Should I file this as a seperate bug?
It seems you have not set up rsync command yet.
As pgcluster uses rsync command to sync db, you need to set it up.
Sorry about that, I figured out what the rsync problem was. A combination of
things really - mismatching rsync versions and a directive in the sshd conf
"AllowGroups" to which postgres was not a member.
I also ran into the problem with a 32-bit build on one server not being able to
function with a 64-bit build on another server (the start_recover fails due to a
checksum error). Is there a workaround for this?
I'm not expert of pgcluster but in my experience pgcluster doesn't work well
between 32bit and 64bit machines. (even if you copy database manually(e.g.
pg_dump/pg_restore), pgcluster will stop..)
You might want to ask it in pgcluster mailing list.
I'm closing this bug for now.
Feel free to reopen if there are any problems yet.