unable to backup cluster client after nt4 - win2K upgrade?

[jtspeed]

Everyone,

Thanks for taking time to reply to this help request.
Our exchange cluster was recently upgraded from NT4.0 to windows 2000. Legato ver 6.11 is running on a separte server and prior to the upgrade was successfully backing up the cluster nightly. After the upgrade the backups began to fail. I have re-installed the client software and I have created new clients. Legato support has been unable to resolve. The only thing that was changed on the cluster was the NICs. There are no errors generated by the jobs and I can push a backup from the client which runs successfully does anyone know of anything that would help me to resolve this problem.

Thanks,

jtspeed

 

[BW78]

What kind of error message occourse?

 

[keefb]

Was there any upgrade of NetWorker Software at all involved ? Regards
KeefB

 

[jtspeed]

The following is the detail received when you right click on the group and click detail.

Unsuccessful savesets
* servername:C: 1 retry attempted
* servername:C:nsrexec authtype nsrexec
* servername:C:type: NSR client description

The backups of virtual clients fail.
The backup of cluster node 1 C:|drive fails
The backup of exchange shares fail.
The only successful backup is of the cluster node 2 files shares and C:|drive.
There are no other errors or alerts to decipher.

 

[goeze]

Did you enter in the backup command field :&quot;save -c <share_name or virt.server> ????

What did you do with the &quot;old&quot; index files after you upgraded from nt to win2k. Possible you have to rename the old client-indexes and create new one for the win2k-servers.

 

[JimTaylor]

Concur with goeze. A procedure exist for renaming clients, if not done, you could be stuck with going to old name and having new name as alias.

Good luck!

 

[redflo]

Hello,

We have a similar Problem:

Backup worked fine, but after some months (10 monts?) one node of the Win2k Cluster failed. All virtual hosts at the clusternode failed too. A reboot of the clusternode fixes the Problem. But some days later the other node fails -> reboot, works -> other node fails ...

a save -vvv somefile on the C:\ stops in the middle of the scan odf C:\ and waits there for the next reboot. So we did EXTENSIVE filesystemchecks. But it doesn't matter if fsck says that C: is OK or not, after some days the node WILL FAIL to backup. This is now for 3 months. Every third day a reboot of a clusternode.

We did all nsrck's mminfo's, nsrim's save's savegrp's, fsck's, sent logfiles of the Networker-Server (Solaris) and Clients to our 1st Level support, but noone knows what to do?

Does someone out there have a hint?

 

[jtspeed]

Thank you everyone who took the time to respond to this thread.
Legato has identified the NIC as the source of the problem. The NIC is an adaptec ANA64044 64-bit / 66 MHz 4 port card. The original card was removed prior to the win2K upgrade because it was not on 2000 HCL. The new card uses a feature called &quot;ScatterGatherDMA&quot;.
I was asked to disable the feature but I could not.
Adaptec has a newer driver than the one win2K installs released October 2001.
Once I upgraded the driver,for all 4 ports on the card, my scheduled backups ran successfully. This was done for both cluster nodes.

Thank You,

JT Speed