jrs
2008-Jun-04 17:53 UTC
[Samba] CTDB problems: 1) Unable to get tcp info for CTDB_CONTROL_TCP_CLIENT, 2) ctdb disable doesn't failover
greetings, trying to follow tridge's failover process at http://samba.org/~tridge/ctdb_movies/node_disable.html I encounter this error. oss02:~ # smbstatus -np Processing section "[homes]" Processing section "[profiles]" Processing section "[users]" Processing section "[groups]" Processing section "[local]" Processing section "[i3]" Unable to get tcp info for CTDB_CONTROL_TCP_CLIENT db_open_ctdb: opened database 'sessionid.tdb' with dbid 0xc0bdde6a Samba version 3.0.26pre1-SVN-build-UNKNOWN PID Username Group Machine ------------------------------------------------------------------- That's odd but I can still run without -np and get some information oss02:/ # smbstatus Processing section "[homes]" Processing section "[profiles]" Processing section "[users]" Processing section "[groups]" Processing section "[local]" Processing section "[i3]" Unable to get tcp info for CTDB_CONTROL_TCP_CLIENT db_open_ctdb: opened database 'sessionid.tdb' with dbid 0xc0bdde6a Samba version 3.0.26pre1-SVN-build-UNKNOWN PID Username Group Machine ------------------------------------------------------------------- Service pid machine Connected at ------------------------------------------------------- db_open_ctdb: opened database 'connections.tdb' with dbid 0x17055d90 i3 1:25892 super-gen01 Wed Jun 4 11:40:00 2008 db_open_ctdb: opened database 'brlock.tdb' with dbid 0x1421fb78 db_open_ctdb: opened database 'locking.tdb' with dbid 0x42fe72c5 No locked files But, when I disable the current node oss02:~ # ctdb disable -n 1 oss02:/ # ctdb status Number of nodes:3 pnn:0 10.200.20.55 OK pnn:1 10.200.20.56 DISABLED (THIS NODE) pnn:2 10.200.20.58 OK Generation:643008338 Size:3 hash:0 lmaster:0 hash:1 lmaster:1 hash:2 lmaster:2 Recovery mode:NORMAL (0) Recovery master:1 the files stop copying, the private interface goes down on node 1 and there is no failover. Eventually, in the screen where I'm running a copy loop, I see: cp: cannot create regular file `ls.4190.exe': No such host or network path `/bin/ls.exe' -> `ls.4191.exe' cp: cannot create regular file `ls.4191.exe': No such host or network path If I reenable the node, the private interface comes up and the files start copying again oss02:~ # ctdb enable -n 1 `/bin/ls.exe' -> `ls.4539.exe' `/bin/ls.exe' -> `ls.4540.exe' My config is: oss02:/ # perl -ne '/^#|^\s*$/ and next;print' /etc/sysconfig/ctdb CTDB_RECOVERY_LOCK="/net/lmd01/space/ctdb.lock" CTDB_PUBLIC_INTERFACE=eth3 CTDB_PUBLIC_ADDRESSES=/etc/ctdb/public_addresses CTDB_LVS_PUBLIC_IP=10.200.180.180 CTDB_NODES=/etc/ctdb/nodes CTDB_DBDIR=/var/ctdb CTDB_DBDIR_PERSISTENT=/var/ctdb/persistent CTDB_LOGFILE=/var/log/ctdb.log CTDB_DEBUGLEVEL=1 oss02:/ # cat /etc/ctdb/public_addresses 10.200.180.181/16 10.200.180.182/16 10.200.180.183/16 oss02:/ # cat /etc/ctdb/nodes 10.200.20.55 10.200.20.56 10.200.20.58 Thanks for any help, JR