| Title: | ase |
| Moderator: | SMURF::GROSSO |
| Created: | Thu Jul 29 1993 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 2114 |
| Total number of notes: | 7347 |
Hello,
I'm clustering four 8400s' at DUNIX 4.0b with TCR 1.4. I configured all
four systems up to the point of adding start and stop scripts on each.
Then I found the ASE database on one had initialized itself because
when I type "asemgr" it asks for a list of server members as it did
when I first started the configuration.
I have tried deleteing this system as a memeber but I get a database
lock error.
I have tried adding it and the primary system thinks it's still part of
the cluster.
I have ftp'd a asecdb from one of the other systems and I can then get
into asemgr but I can't do anthing because the aseagent is in setup
mode with the -p switch on.
I have tried restarting the agent but it goes right back into setup.
Is there anything I can do to get it back to the configuration state
it was previous? I have no idea how the original database got blown
away and I don't have a backup of it.
Do I have to reconfigure the cluster from scratch????
Thank you in advance
Dennis
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1888.1 | MSBCS::GRADY | Wed Feb 19 1997 15:48 | 16 | ||
As further info I went ahead and reconfigured all nodes from scratch brought everything back to where it was and lost it a second time. Noone touched or was logged into the cluster except me. I tried running asemgr and the same node came up and asked me for a kist of servers etc.... Any help would really be appreciated, Thanks in advance Dennis | |||||
| 1888.2 | Some questions | BRSDVP::DEVOS | Manu Devos DEC/SI Brussels 856-7539 | Wed Feb 19 1997 16:31 | 19 |
Hi,
After you add the failing node to the cluster from the first system,
can you see the status of this added failing node with asemgr -dh.
I have the feeling that the communication/hostname with the failing
node are not OK.
So, if you add node B, C and D from asemgr on node A, what is the
result of asemgr -dh from node A just after?
Are you able to ping node D from node A ?
Are the permission associated with /var/ase/config/asecdb ok ?
What is telling you the daemon.log and kern.log on node A and node D?
Regards, Manu.
| |||||
| 1888.3 | MSBCS::GRADY | Thu Feb 20 1997 10:47 | 33 | ||
In answer to your first question the first node thinks the failing node is still configured in the cluster. The failing node has no information in it and asks for a list of servers. I am able to ping the failing node. The permissions on the asecdb file are the same as on all other nodes. The daemon log come up with a database lock error and that's the only reference time was to it. The kernel.log tells me nothing. My major concern is yesterday the asecdb file became corrupt or something because when I tried to get into asemgr it came up again with wanting a list of the servers. when I reconfigured the system a second time I made a copt of asecdb in /usr/var/ase/config, I copied the backup to asecdb and rebooted the machine and it came backup fine. Dennis | |||||