TSM reclamation fails with damaged file on tape

[Milan Stubniak]

Hi ,

we are testing quadstor-vtl 3.0.16 used by our TSM server 8.1.0. We have virtual library IBM 3584 Ultra Scalable Library. Everythink is  fine, except reclamation of tape. Usually reclamation fails with damaged files on tape. (

ANR1161W The move process is skipping a damaged file on volume B00009L5) and tape goes read-only.  

In TSM activity.log we have also :

ANR8311E An I/O error occurred while accessing drive DRV31 (/dev/rmt-vtl11) for

READ operation, errno = 78, rc = 2863.

 

ANR9999D_0252090925 NtpValidateComBlockHdr(pvrntp.c:6100) Thread<964076>:

Invalid block header read from NTP drive DRV31 (/dev/rmt-vtl11).

(magic=5A4D4E50, ver=5, Hdr blk=2464348 <expected 2464347>, dbytes=262096

<262096>)

In VTL server /var/log/messages we have 

kernel: WARN: __tdrive_cmd_validate_write:2265 retval 17

Is it missconfiguration or a bug? Should I use another type of library or drives? 

Any idea what can i do?

Thanks for response.

 

[QUADStor VTL Support]

The warning in the kernel log messages is ok.

Is this with every vcartridge or only with B00009L5 ?

Is this an occurance with 3.0.16 but not with the earlier versions ?

Does an AUDIT VOLUME for B00009L5 go through fine ? If yes, then its
possible that a space or locate isn't correctly positioning at the
required block.

> --
> You received this message because you are subscribed to the Google Groups
> "QUADStor VTL" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to quadstor-vtl...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Milan Stubniak]

No the problem is not only for B00009L5, after 4.4.2017 this problems apperars on every tape in reclamation process.

Probably yes , before 3.0.16 we haven`t these problems. I think 14.3.2017 we realized upgrade quadstor to 3.0.16 and also we create another VTL library.  We noticed this problem in 4.4.2017. (Probably first occurence).

Audit volume go through , but is it ok? Can i fix it somehow?

[QUADStor VTL Support]

What was the version installed prior to 3.0.16 ?

kernel: WARN: __tdrive_cmd_validate_write:2265 retval 17

The above warning maybe of importance, unlike we first thought.

We don't know yet if the issue is while writing or reading. If its
while reading then its easily fixable.

[Milan Stubniak]

Hi,

problem is finally resolved. It was caused by HW issue - one of two FC port VTL servers is definitely broken. After some setup (delete every FC zone to port - even storage ), it works without issue.

No qla scsi reset in VTL logs, (no I/O erros in TSM). Backup,restore works as expected.

Thank you for your quick response and help.

Regards,

M.