Tape Libraries and Drives
cancel
Showing results for 
Search instead for 
Did you mean: 

Tape Libraries going offline periodaclly

Ayman Altounji
Valued Contributor

Tape Libraries going offline periodaclly

I have a MSL5026 Connected to a MDR,this is connected to 5 servers running WINDOWS 2000with a 8 port switch.I have a problem were every so often windows picks up that the MDR has been removed and the Library in Backupexec 8.6 says the the tapes are offline.I have replaced the MDR, the Scsi ports in the MSL and also the fibre cable and gbics.i also have the latest drivers for the fibre card.
has anyone got and idea.
6 REPLIES
Ayman Altounji
Valued Contributor

Re: Tape Libraries going offline periodaclly

1. On the Devices Tab in Backup Exec, right click on the Tape Drive(s), then select the Properties, Configuration Tab. Uncheck "Enable Device for Backup Exec", Apply, OK. The devices will show disabled after doing this. Right click on them again and select "Delete". Do the same for the loader that shows under Robotic Libraries until all the devices are gone. Make sure that nothing shows under Stand-alone Drives also.

2. Set all of the Backup Exec Services to manual. Stop/disable the Removeable Storage Service.

3. Shut the server all the way down. Cycle power to the Robotic Library until it is at an Online, Ready State, then power up the controller (if it is a Fiber Controller) fully prior to Cold-Booting the server.

4. After logging on to the server, give the OS/Applications at least 5 minutes to fully load prior to starting the Backup Exec Services.

5. Launch Backup Exec and run backups to test.
Ayman Altounji
Valued Contributor

Re: Tape Libraries going offline periodaclly

Please, send me information of version of firmware All hardware that you used in this case

Braulio Sanchez
braulio.sanchez@gzeos.com

Re: Tape Libraries going offline periodaclly

I also have this problem except I use an MSL5026SL connected to a Network Storage Router N1200. Every couple of days the library will appear as offline in Backup exec 8.6 (build 3878) and backups go into a Paused status. This is a real pain because no errors are logged in the W2K event Viewer or anything. My N1200R connects to a 2GB SAN switch 8/EL and everything is at the latest firmware. I feel like ripping the whole lot out and going back to a non-shared SCSI solution. If anyone can suggest a solution to this I would be VERY grateful.
Eric Pero
Occasional Advisor

Re: Tape Libraries going offline periodaclly

I'm experiencing the same problem. I have a ML370 cluster with Compaq Storage router (2404) connected to a MSL5026 library. I just updated the firmware on the MSL5026 (now shows HP logo and menu on it :) and I am getting drive offline via Backup Exec 8.6. The other node operates fine and sees the library and drives. I thought it was the SDLT tape/loader drivers, so I replaced them, but that didn't help. If I cold boot the server, storage router, and library, it will be fine for a while, then it just loses the drive. Anyone have any ideas? (I tried removing all of the devices and having Win2K re-add them and then have Backup Exec re-add them, but that didn't work either).

Thanks,
Eric Pero
The Hartford
Intel Server Team
Ross Humphryes
Frequent Advisor

Re: Tape Libraries going offline periodaclly

Hi, I am sure this isn't a Backup Exec issue as we have the same problem and we use Legato NetWorker 6.2.
Basically we first notice the problem when NetWorker cannot load or move the tapes. There is a command called INQUIRE.EXE that we run which scans the scsi bus and it shows us that the tape devices are gone. We also have to reboot the Library and the MDR to get the server to locate the tape devices again and sometimes the server also needs to be rebooted. This happens every week.
Get in my belly
Vladimir Vybiral
Valued Contributor

Re: Tape Libraries going offline periodaclly

ooops, some of the things happening to some of You look as there are some LIPs (loop initialization process) coming on Your fibre - if this happens during some open SCSI command on fibre (SCSI reset, cluster failover, device adding or removal or reset), which is set to loop topology, the backup tapes are sometimes also lost, backups interrupted etc. Make sure You use switched fabric topology, or some sort of LIP control, or quickloop on brocade switches. Also, use hard_alpa addressing on the device (tape library). Hope some of this helps, BR, Vladimir
When speaking, Your words should sound better than Your silence - Arabic proverb