Tape Drive Going Offline

[AgeasServer]

Hi,

We have a ongoing issue with our backup exec software where a tape drive keeps going offline, currently using BackupExec 2010 R3, it seems to only occur when a certain amount of data is transferred, checking the logs on the server there is no errors in the event viewer, looked at the logs on the fibre switch and there are no errors here as well.

To resolve the issue i have to reboot the tape library and then restart the backup exec services. The drivers that the tape drives are using are Symantec Drivers. The drive drops out of device manager as well and cant see why it is doing this.

It is a IBM TS3200 tape library, does anyone have any suggestions, I have run IBM update xpress to make sure all drivers on the server are up to date.

Thank You

[VJware]

If the drive is not visible in the Dev Mgr, then Backup Exec would not recognize it either. Check the Event Viewer - System logs for any hardware errors (typically events 5, 7, 9, 11, 15)

[CraigV]

Hi,

Seems to be some sort of hardware-related problem. IBM have a diagnostic utility you can download and run against the tape library. This is going to show any possibly hardware issues, so I'd suggest you do this.

Also make sure that the FC HBAs in the server have had their drivers updated.

Thanks!

[AgeasServer]

The first event i receive is this

[9720] 04/15/13 05:15:30 Adamm Mover Error: DeviceIo: 04:00:00:00 - Device error 1117 on "\\.\Tape0", SCSI cmd 0a, 1 total errors

%2

%3

%4

%5

%6

%7

%8

They all appear under the event ID 33152

Thanks

[AgeasServer]

Craig, thanks for the response, I have updated the Fibre HBA card drivers and have also put the IBM tape drives diagnostic utility on the server it shows no errors, but also it cant interrogate the drive when it disappears out of the device manager.

However, when i go to the web interface page for the tape library all drives are showing as ready and i can manually move the tapes in and out of the I/O or drives

[AgeasServer]

Craig, thanks for the response, I have updated the Fibre HBA card drivers and have also put the IBM tape drives diagnostic utility on the server it shows no errors, but also it cant interrogate the drive when it disappears out of the device manager.

However, when i go to the web interface page for the tape library all drives are showing as ready and i can manually move the tapes in and out of the I/O or drives

[CraigV]

...that would happen if the drive dropped out of Windows. Have you tried to run the diagnostics when the drive is online in Windows? Have you tried to use another FC card?

Thanks!

[AgeasServer]

Yeah, when the drive is online and visible in windows it works fine and all test are passed.

What i have done as a test is to remove this drive from the equation, disabling it in backup exec and allowing a full back up to run using the other drive. If it goes offline again then I'd imagine its not that drive that is causing the issue. If it doesnt go offline then i'd imagine its the drive or the hardware connected to it.

Might be worth if it is the latter to swap the paths over and see what happens, will let you know, thank you for your update.

Thank You

[Larry_Fine]

it seems to only occur when a certain amount of data is transferred

Can you expand on this point?  How much data?  How much run time?  Might you have some sort of a heating issue that after running long enough, there is some thermal event in the library or tape drive?

[AgeasServer]

I have run a test backup overnight using the over drive and once again the library has gone offline in backup exec. This is not the same drive, so therefore it is not using the same fibre card, not using the same cables and not using the same drive. Any suggestions?

[AgeasServer]

Of course, there is no specific amount of data sometimes it can be as little as 100GB and sometimes it can be as much as 1.5TB it never really goes above the maximum there, I am told that this started happening after a upgrade of Service Pack to BE 2010 but i cant see that being a direct impact on it.

Do you know where i might be able to find thermal events on the library itself and i will check, i did run the IBM drive diagnostic tool and this did say that everything was fine but this probably only examines the drive at that point

[Dharamveer_Yada]

Have you reinstall the symantec tape drive ??

[AgeasServer]

I have uninstalled from device manager the LTO5 Tape Library, then gone into backup exec and gone to wizards to install the tape library, I have also done a full reinstall of backup exec, this was a full uninstall not just an application uninstall

[CraigV]

...log onto the web front-end of the library and take a look there. It would have an event viewer that you can look at to see if the thermal events show up.

Thanks!

[CraigV]

...get the drives online, stop the BE services and run an in-depth IBM diagnostic check on the library.

The other option is to completely remove it from within BE and Windows, and restart the server. Once done, make sure that the robotics (medium changer) shows up as Unknown Medium Changer in Windows Device Manager, and then use tapeinst.exe to install the BE drivers for the drives.

Have you also ensured there are no FC switch port issues?

[lmosla]

Have you done a full power cycle on the tape drive by performing the following steps?

1. Power Off the Server completly
2. Power off the Tape Drive completly
3. Power on the Tape Drive wait till it is at a 'Ready' State and then
4. Power on the Server

Thanks