I had a box hard lock, everything was connected, happy, etc. After rebooting it
i brought back up the ndb processes which came up fine. i then went to start
back up the sql server and started getting the following:
2006-04-20 09:30:10 [MgmSrvr] WARNING -- Allocate nodeid (0) failed.
Connection from ip 12.34.56.78. Returned error string "Connection done
from wrong host ip 12.34.56.78."
2006-04-20 09:30:10 [MgmSrvr] INFO -- Mgmt server state: node id's
2 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 connected but not
reserved
2006-04-20 09:30:10 [MgmSrvr] INFO -- Mgmt server state: node id's
7 not connected but reserved
i've changed the ip address, but not the fact that it is the same at both
occurrences. the management server seems to think the ip address don't match.
i've double checked the config.ini and my.cnf and the node ids were correct
(and worked for a couple weeks before the host died.)
i'm seeing the above error message in the mgm logs about every 10 or 15
seconds. which is about the time it takes the mysql process to come up, die b/c
it can't support the default table type of ndbcluster, and try again. it was
filling up the mgm server logs. there isn't anything useful interesting in the
mysql logs, just it trying to start up and failing with:
060420 9:30:49 [ERROR] Default storage engine (ndbcluster) is not available
i'm guessing that restarting the mgm nodes will make things better, which is
the next thing i'm going to try. anyone else seen this before?
--
-rm
