Start a Conversation

Unsolved

This post is more than 5 years old

5832

August 26th, 2010 12:00

Periodic tape backup throughput issues

I’m having a perplexing throughput issue with my tape backups.  Every two weeks (on a Tuesday for the past two months) my tape backup throughput decreases dramatically – from 1,360 MB/min to 130 MB/min.  The speed decrease happens regardless of the source of the backup – over the LAN or running a duplicate job where the backup-to-disk folder is stored on the backup server hard drive.  This situation does not impact the speeds of disk-to-disk backups.  The only way to resolve the issue is to reboot the backup server.  Below is my configuration.

Dell PowerEdge 2850

Dell PowerVault 124T LTO-3

Adaptec 39160 SCSI adapter

Windows Server 2003 Std x64 SP2 fully patched

Backup Exec 2010 fully patched

Troubleshooting steps I’ve taken

·         contacted Symantec who had me run various diags

·         contacted Dell who had me upgrade the firmware for the tape/autoloader

·         checked to make sure there are no scheduled tasks running during this time

·         checked Task Manager for processing consuming CPU or generating lots of I/O – none found

·         compared running processes to server after reboot – no differences

·         rebooted server for several consecutive days to see if it would extend the interval – it did not

·         verified all the drivers were at the latest levels – tape, autoloader, SCSI adapter

·         checked Windows Application and System logs for errors – none found

·         checked tape autoloader logs for errors – none found

·         checked tapes for errors – none found

Anyone have any ideas what could be causing those or what other steps I could take to troubleshoot the problem?

899 Posts

August 27th, 2010 09:00

Very odd situation..did this just start happening in the last 2 months or has it always been like this?

 When you reboot the server to correct the problem are you also powercycling the 124T or does it continue to stay powered on?

Did either tech support team run any sort of throughput tests w/ you while the server was in the "slow" state? If not I would recommend running a test on the drive w/ an external tool, get backup exec  and your file system out of the picture. If the unit has a Quantum drive you will need Xtalk, if it is an IBM drive you will need ITDT, both of these should be on the Dell support site under the 124T. Be sure when you run the tool the Backup Exec services are turned off. Get a baseline when the drive is running correctly wait your two weeks or until Tuesday  then when it starts acting up again run the tool again. Depending on the outcome it may give you a path to follow to determine the bottle neck....

 

you got the gremlins in on this one for sure it'll be interesting to see how this gets solved.


Bob

 

 

 

 

 

 

September 3rd, 2010 08:00

This has been happening pretty consistently since the beginning of March.  Dell support has run ITDT during the slowdown times.  It is currently in this state since last night.  Very frustrating.

No Events found!

Top