Unable to connect to remote agent

[poktor]

Hi,

For a few days now, i keep getting failed backups and this in de BE Debug Monitor:

BENGINE:  [10-18-11 16:56:43] [0000]     [1824] 10/18/11 16:56:43 Failed to open Microsoft cluster ()
BENGINE:  [10-18-11 16:56:43] [0000]     [1824] 10/18/11 16:56:43 VCS cluster keys do not appear to be present in the registry
BENGINE:  [10-18-11 16:56:43] [0000]     [1824] 10/18/11 16:56:43 Failed to open VCS cluster ()
BENGINE:  [10-18-11 16:57:26] [0000]     BECryptoInit: BECrypto non-FIPS mode successfully enabled.
BENGINE:  [10-18-11 16:58:08] [0000]     BECryptoInit: BECrypto non-FIPS mode successfully enabled.
BESERVER: [10-18-11 16:58:09] [0000]     [5320] 10/18/11 16:58:09 "Cluster" key does not appear to be present in the registry
BESERVER: [10-18-11 16:58:09] [0000]     [5320] 10/18/11 16:58:09 Failed to open Microsoft cluster ()
BESERVER: [10-18-11 16:58:09] [0000]     [5320] 10/18/11 16:58:09 VCS cluster keys do not appear to be present in the registry
BESERVER: [10-18-11 16:58:09] [0000]     [5320] 10/18/11 16:58:09 Failed to open VCS cluster ()
BKUPEXEC: [10-18-11 16:58:10] [0000]     UIVirtualObjectView(JobHistoryView) Queried 500 items: 0 to 499
BKUPEXEC: [10-18-11 16:58:23] [0000]     UIVirtualObjectView(JobHistoryView) Queried 500 items: 0 to 499

This is all the remote agent has logged:

"<<<<<<<<<<<< LOGGING ON >>>>>>>>>>>>>>>>>>>"

when i try a restore of a mailbox to a recovery storage group, i get this:

"The job failed with the following error: Unable to attach to a resource.  Make sure that all selected resources exist and are online, and then try again.  If the server or resource no longer exists, remove it from the selection list. Edit the selection list properties, click the View Selection Details tab, and then remove the resource."

I already searched the web and this forum, but all proposed solutions i found until now were unsuccessful... :(

I also checked for the mentioned cluster key, it does exist. Permissions full control for local system and domain admins.

The backup server connects to an exchange 2007 cluster of 2 nodes. Backup software = BE12.5. Remote agents already removed and reinstalled. Licenses removed and reinstalled.

Any ideas, anyone?

Thanks in advance.

Cheers,

Richard

edit: log uploaded.

[newsolutionBE]

Hi

Is this exchange 2007 sp3 if yes then BE 12.5 doesn't support exchange 2007 sp3 so please upgrade this to BE 2010 r3 sp1.please check SCL below

http://www.symantec.com/docs/TECH62301       for12.5

BE 2010 R3 SP1 - http://www.symantec.com/docs/TECH169846

BE 2010 R3  Installer - https://www4.symantec.com/Vrt/offer?a_id=93355

Also the log shown by you above is of no help if you want to debug you can enable beremote agent debuging on both the nodes of cluster

for same you can open registry & then go to hklm-software-symantec-backupexec for windows-backupexec-engine- logging --in which you will see enable debug key to 1 from 0 do same on both cluster node & also on media server

Also please upload the job log error or job log so that can help you

Also in end please check this document too if relevant to your enviorment

http://www.symantec.com/docs/TECH66459

Thanks

[poktor]

@newsolutionBE: it is EX2007, but not with SP3. I did as you asked and enabled debug logging on all machines. I replaced the log in my first post. What you see now, is all there is. What you see now, is the result of 1 restore-job i started.

Like i said, everything is welcome, this is already taking 3 days, which is way too long..

cheers,

Richard

[CraigV]

Hi Richard,

You get that message mainly when the RAWS agent loses connectivity to the media server. Now...since you're backing up a cluster, have you selected the individual nodes to be backed up, or the virtual node? It should be the virtual node...and to restore, you'd point to the virtual node too.

Is your media server patched with the latest patches available from the Symantec site, and have these patches been push-installed to your remote servers?

Thanks!

[poktor]

Hi Craig, Yes, we connect to the virtual node. Tried the individual nodesof course now, but to no avail.. we have not yet looked at the patches, will do hat tomorrow.

[newsolutionBE]

Hi

Can you please post the job log as log posted by you are of no help also if you could let me know if you have enabled host file entry on exchange server if it is on win2008

Thanks

[poktor]

Hi,

Log is now posted, hope it's of any use to you. EX2007 is on win2008, but IPV6 is disabled on the network adapters of the exchange servers.

Thanks for helping.

Cheers,

Richard

[poktor]

update: i installed BackupExec SP4. Update of the remote agents went fine and without problems. This SP improved things a bit, although backup and restore is not yet possible. Errors are now limited to "communications error". I added the log to this update.

[CraigV]

Mmm...can you verify whether or not you have a firewall running on the physical nodes? If so, disable that and try running a backup/restore again. If it works, then either leave the firewall disabled, or put in an exclusion for the RAWS agent.

Has this worked and then become a problem, or never worked?

[poktor]

It all worked fine, until last week. All servers are in the same subnet, no firewall. Windows firewall  is off on all machines. I just noticed something strange: the common node shows nothing: no folders, no mailboxes, nothing. It's empty. I think this is not good, is it? Or is BE incapable of backing up the common node and must it be pointed to the individual nodes?

One more thing: i also get a message "Microsoft information store not found". when i use google on this, i get a lot of questions, but no answers. Is there someone in here who can point me in the right direction?

[CraigV · Accepted solution]

Hi,

The reason for backing up the virtual node is that BE doesn't get affected when the physical nodes take ownership from each other when they fail over.

Try backing up the current active physical node and see what happens.Also check the Event Viewer logs and make sure you don't have any issues with that cluster...