cancel
Showing results for 
Search instead for 
Did you mean: 

Status 84 - crc error - with VTL

bbahnmiller
Level 4

Has anyone run across status 84 crc errors on a VTL? We had a flurry of about 9 status 84 errors this weekend. The errors ran across 5 different media servers, 2 different VTL's (EMC DL3D 3000's), 9 different VTL drives, 6 different clients running Unix, Windows and NDMP backups.

The only significant factor I see is that they mostly occurred with 14 minutes. The 2 ndmp backup failures occurred 2 1/2 hours later.

The only troubleshooting step that could possibly relate to a VTL would be corrupt drivers. Given how several hundred other backups ran just fine on those same media servers, I don't see how that is the case.

The other thing is that these failures occurred during peak backup times. Will the VTL's return status 84's if they get overloaded? I'm a bit baffled on this one.

        Bryan

10 REPLIES 10

Nicolai
Moderator
Moderator
Partner    VIP   
The only root cause I can imagine for a status code 84 on a VTL would be if it ran out of bac end disk space.

Did the tape drives return some sort of SCSI sense key ?

bbahnmiller
Level 4

 

The only root cause I can imagine for a status code 84 on a VTL would be if it ran out of bac end disk space.

Nope. Not out of disk space.

Did the tape drives return some sort of SCSI sense key ?

I haven't gotten into any low level logs yet. I'll check the server info and see if I see anything there.

       Thanks!

Nicolai
Moderator
Moderator
Partner    VIP   
Also check the bptm log if enabled. If you can find a sense key from the virtual device, it will help you determine if it was a "real" HW error or a device driver issue.

Sriram
Level 6
Did you investigated by running a backup job at different time, just to make yourself comfortable on isolating the issue further down.

I used to get these errors on my VTL regulary but not all the times.  This went on for more than a week.  Then i decided to remove the drive from VTL and reconfigure it and delete the VTL tapes involved in error 84 and reallocate them too. 

The problem didn't recour form since 2 months.

zippy
Level 6
 replace the drive

zippy
Level 6
 opps rebuild the drive

Rakesh_Khandelw
Level 6
You may want to analyze the physical path, check if all of these affected drives are configured through same Fiber port on VTL or SAN switch. It could be because of bad port on VTL or SAN swithch or a bad cable.

Kevin_Kelly
Level 0
We are getting the same 84 errors implimenting NetApp VTL's.
The 84 errors are sporatic mostly coming on heavy load.
The media servers are also sporatic mix of OS's
The data beeing backed up is also sporatic could be flat file or DB.

Vitals
NBU 6.5.2.a
45 Media servers mix of OS's
VTL's are NetAPP 1400's
Fibre switches Brocade 48000 and DCX's

Here is a snippet from the messages log on a solaris media server during the time the 84 error was reported in the bptm log.

Oct  3 06:32:20 dba09 lpfc: [ID 728700 kern.warning] WARNING: lpfc3:0754:FPe:SCSI timeout Data: x0 x3 x10 x1D3
Oct  3 06:32:20 dba09 scsi: [ID 107833 kern.warning] WARNING: /pci@1e,600000/fibre-channel@3/st@3,1 (st43):
Oct  3 06:32:20 dba09     SCSI transport failed: reason 'tran_err': giving up
Oct  3 06:34:05 dba09 lpfc: [ID 728700 kern.warning] WARNING: lpfc3:0754:FPe:SCSI timeout Data: x0 x3 x10 x35B
Oct  3 06:34:05 dba09 scsi: [ID 107833 kern.warning] WARNING: /pci@1e,600000/fibre-channel@3/st@3,3 (st45):
Oct  3 06:34:05 dba09     SCSI transport failed: reason 'tran_err': giving up
Oct  3 06:36:43 dba09 bptm[11746]: [ID 832037 daemon.error] scsi command failed, may be timeout, scsi_pkt.us_reason = 3
Oct  3 06:38:17 dba09 bptm[13481]: [ID 832037 daemon.error] scsi command failed, may be timeout, scsi_pkt.us_reason = 3
Oct  3 06:47:23 dba09 lpfc: [ID 728700 kern.warning] WARNING: lpfc3:0754:FPe:SCSI timeout Data: x0 x3 x10 x5B5
Oct  3 06:47:23 dba09 scsi: [ID 107833 kern.warning] WARNING: /pci@1e,600000/fibre-channel@3/st@3,4 (st46):
Oct  3 06:47:23 dba09     SCSI transport failed: reason 'tran_err': giving up





Nicolai
Moderator
Moderator
Partner    VIP   

Timeout should not occur unless a device drop off the SAN. Sound like buggy VTL code to me.

zippy
Level 6
 the log say that the drive is not there.  The FC card is timing out, like Nicolai said VTL vendor patch action or the driver for the FC card, compatibility problems....