ACN5798E MS Exchange Error

[jnelson]

We have been running backups of our Exchange 2007 environment using Data Protection of Exchange Agent without issue.
Then...
We moved our disk storage pool from local disk to SAN disk.
Backup speeds improved greatly. Yippie!
But...
Since then we've been seeing ACN5798E error messages (see log below).
The storage pool was moved on 1/28.
We have gotten successful backup since the move, just not consistently.

01/30/2009 19:00:00 ============== Log file pruned using log retention period of 30 day(s)
01/30/2009 19:00:00 ============== No log entries pruned
01/30/2009 19:00:01 =========================================================================
01/30/2009 19:00:01 Request : Backup
01/30/2009 19:00:01 SG List : *
01/30/2009 19:00:01 Backup Type : FULL
01/30/2009 19:00:01 Database Name :
01/30/2009 19:00:01 Buffers : 3
01/30/2009 19:00:01 Buffersize : 1024
01/30/2009 19:00:01 Exchange Server : MAIL02
01/30/2009 19:00:01 TSM Node Name :
01/30/2009 19:00:01 TSM Options File : dsm.opt
01/30/2009 19:00:01 Mount Wait : Yes
01/30/2009 19:00:01 Quiet : No
01/30/2009 19:00:01 -------------------------------------------------------------------------
01/30/2009 20:07:03 Total storage groups requested for backup: 8
01/30/2009 20:07:03 Total storage groups backed up: 8
01/30/2009 20:07:03 Total storage groups expired: 0
01/30/2009 20:07:03 Total storage groups excluded: 0
01/30/2009 20:07:03 Throughput rate: 26,024.87 Kb/Sec
01/30/2009 20:07:03 Total bytes transferred: 107,093,636,176
01/30/2009 20:07:03 Elapsed processing time: 4,018.60 Secs
01/31/2009 19:00:01 ============== Log file pruned using log retention period of 30 day(s)
01/31/2009 19:00:01 ============== No log entries pruned
01/31/2009 19:00:02 =========================================================================
01/31/2009 19:00:02 Request : Backup
01/31/2009 19:00:02 SG List : *
01/31/2009 19:00:02 Backup Type : FULL
01/31/2009 19:00:02 Database Name :
01/31/2009 19:00:02 Buffers : 3
01/31/2009 19:00:02 Buffersize : 1024
01/31/2009 19:00:02 Exchange Server : MAIL02
01/31/2009 19:00:02 TSM Node Name :
01/31/2009 19:00:02 TSM Options File : dsm.opt
01/31/2009 19:00:02 Mount Wait : Yes
01/31/2009 19:00:02 Quiet : No
01/31/2009 19:00:02 -------------------------------------------------------------------------
01/31/2009 19:51:48 Backup of storage group SG7 failed.
01/31/2009 19:51:48 ACN5798E MS Exchange API HRESEBACKUPREADFILE() failed with HRESULT: 0xc7ff07dd -A communications error occurred while attempting to perform a local backup.
01/31/2009 19:51:48 Backup of storage group SG1 failed.
01/31/2009 19:51:48 ACN5798E MS Exchange API HRESEBACKUPSETUP() failed with HRESULT: 0xc7ff0fa4 -Functions called in an invalid sequence.
01/31/2009 19:51:48 Backup of storage group SG2 failed.
01/31/2009 19:51:48 ACN5798E MS Exchange API HRESEBACKUPSETUP() failed with HRESULT: 0xc7ff0fa4 -Functions called in an invalid sequence.
01/31/2009 19:51:48 Backup of storage group SG3 failed.
01/31/2009 19:51:48 ACN5798E MS Exchange API HRESEBACKUPSETUP() failed with HRESULT: 0xc7ff0fa4 -Functions called in an invalid sequence.
01/31/2009 19:51:48 Backup of storage group SG4 failed.
01/31/2009 19:51:48 ACN5798E MS Exchange API HRESEBACKUPSETUP() failed with HRESULT: 0xc7ff0fa4 -Functions called in an invalid sequence.
01/31/2009 19:51:50 Retrying failed backups...
01/31/2009 19:51:51 Total storage groups requested for backup: 8
01/31/2009 19:51:51 Total storage groups backed up: 3
01/31/2009 19:51:51 Total storage groups expired: 0
01/31/2009 19:51:51 Total storage groups excluded: 0
01/31/2009 19:51:51 Throughput rate: 13,444.37 Kb/Sec
01/31/2009 19:51:51 Total bytes transferred: 42,730,599,330
01/31/2009 19:51:51 Elapsed processing time: 3,103.84 Secs

We've searched the web and ADSM.org - but no help.
Any ideas anyone?

 

[moon-buddy]

What version of TDP are you using? Have you tried reinstalling the BA and TDP clients?

 

[jnelson]

TDP verison is 5.5.1.0.

I Hadn't considered reinstalling the TSM Clients. How would that make it better when all we did was change the Storage Pool. Our Exchange Admins will make a fuss if I don't have some logic behind that request.

I have seen a few tech docs that suggest we reboot the exchange servers.

 

[jnelson]

Because the exchange backup is causing a cluster failover, we are trying to switch our tcpclientaddress to the local IP instead of the cluster IP.

The exchange backups are initiated via command line.
The dsm.opt file location is specificed in the tdpcfg file.
We've made the change in the dsm.opt file but have not seen the change on the backups.
In dsm.opt we added
tcpclienta 172.10.10.10

Unsure why

Do I need to update the node on the server side?

Attached is some additional query information
ACN5057I The C:\Program Files\Tivoli\TSM\TDPExchange\tdpexc.log log file has been pruned successfully.
Data Protection for Exchange Preferences
----------------------------------------
BACKUPDESTination................... TSM
BACKUPMETHod........................ LEGACY
BUFFers ............................ 3
BUFFERSIze ......................... 1024
DATEformat ......................... 1
LANGuage ........................... ENU
LOCALDSMAgentnode...................
LOGFile ............................ tdpexc.log
LOGPrune ........................... 60
MOUNTWait .......................... Yes
NUMberformat ....................... 1
REMOTEDSMAgentnode..................
RETRies............................. 4
TEMPLOGRESTOREPath..................
TIMEformat ......................... 1

C:\Program Files\Tivoli\TSM\TDPExchange> tdpexcc query tsm
IBM Tivoli Storage Manager for Mail:
Data Protection for Microsoft Exchange Server
Version 5, Release 5, Level 1.0
Copyright IBM Corporation 1998, 2008. All rights reserved.
Tivoli Storage Manager Server Connection Information
----------------------------------------------------
Nodename ............................... MAIL03_TDP
NetWork Host Name of Server ............ 196.20.20.20
TSM API Version ........................ Version 5, Release 5, Level 1.0
Server Name ............................ TSM Server
Server Type ............................ AIX-RS/6000
Server Version ......................... Version 5, Release 5, Level 1.1
Compression Mode ....................... Client Determined
Domain Name ............................ MAIL_DOMAIN
Active Policy Set ...................... STANDARD
Default Management Class ............... MAIL_MANAGEMENT_CLASS

 

[jnelson]

The saga continues...
Apparently we can not set the tcpclientaddress using client initiated backups. The client will use the IP path of least resistance. So we are looking at using our firewall to block the traffic on the public NIC to teh TSM server so it will use the other private NIC for backup.

We are guessing that the faster backups are saturating the network causing the cluster to failover.

We are also looking at lowering the buffers to 2 which will make the backups take longer but won't saturate the NICs\network.

Pressures on......

 

[jnelson]

One other nice tidbit from IBM...
They said they don't support MS Exchange 2007 Clustering.

 

[spiffy]

sorry to hijack this, but IBM said they do not support TSM on an exchange 2007 cluster???
WTF?? is this documented somewhere?
This may put a dent in our new exchange cluster that is on the books.....

 

[N.joshi]

Hey...
Ne solution for this problem......
facing same......



Njoshi