ANR8302E Errors with HP-LTO3 Drives

[Mikey D]

Guys having some real struggles here with this and the vendors aren't able to track down what's causing this yet.

Setup is a Solaris 8 V890 running TSM 5.3.6.0 with four dual port 2G Emulex LP10000 HBA's. We have twenty HP-LTO3 drives for this machine. I know a lot of unsupported versions OS and TSM wise, we've got a server we're going to refresh to but it can't happen soon enough unfortunately and these errors are causing us some real problems just keeping scratch available and trying to run any sort of migrations, reclaims, etc...

For some reason we take these errors below, happens on different media, & drives every few hours. Hoping someone has run into something similar in their environment. Absolutely has me mystified on what is causing this. We've tried downgrading drive firmware, upgrading drive firmware to no avail. HBA firmware is up to date and so is everything else as much as can be on a Solaris 8 platform.

08/17/09 13:12:01 ANR8302E I/O error on drive SL8500 (/dev/rmt/6mt) with volume R01156 (OP=LOCATE, Error Number=201, CC=412, KEY=02, ASC=3A, ASCQ=00, SENSE=70.00.02.00.00.00.00.10.0- 0.00.00.00.3A.00.00.00.94.50.00.00.00.00., Description=M- edia not present in drive). Refer to Appendix D in the 'Messages' manual for recommended action. (SESSION: 14664, PROCESS: 245)

[picay]

what driver did you use ?

[Cheeky]

What library are the drives in?

We had some similar errors on the ESL712e and after the latest firmware upgrade about 6 months ago we get the odd error but nothing like what we used to.

[Mikey D]

Driver hba wise is the Emulex LPFC 6.10g version. TSM MT driver for the HP LTO-3 tape drives.

Firmware wise on the drives we're running L6CS for the 2Gb attach drives and M66S on the 4Gb attach drives. This is in a SUN/STK SL8500 silo.

[picay]

it remind me a very strong problem with firmware I had once with LTO3 (IBM flavour)
At system level, did you have some message/logs ?
Check with hba driver if it is possible to enable a trace and try to reproduce the problem

[picay]

a SL8500 silo mmmm nice !!!!

[SleeperGTP]

Did you enable encryption at any time?

We purchased an HP ESL 712e to start on a new implementation of TSM to replace our aging equipment. The libraries were extremely difficult to implement due to lack of documentation and random issues. During the implementation we enabled encryption on one of the sequential storage pools and started receiving the same problem as you. After 4 months of working with IBM they released a new version of TSM which was 5.5.1.1 which fixed the issue of drive encryption. When an ecrypted tape was inserted, there was a bit in the driver that was changed to a 1 but not reset upon dismount. If you rebooted the library/drive the issue would go away for a while. I can help you out a lot on this issue.

[Mikey D]

Can't say we've enabled encryption at all on these drives, it was set to allow in the device class originally which I did change to no just to be safe. None of the storage pools however had encryption enabled. Getting to support levels is the main part of our problem, I've got a T5220 Solaris 10 server with TSM 5.5.2.1 loaded ready to migrate to but getting the window to do it is the sticking point right now.

[SleeperGTP]

Are you getting any ANR877W's? Have you replaced any drives or added new tapes to the library?

[Mikey D]

Are you getting any ANR877W's? Have you replaced any drives or added new tapes to the library?

We had some problem drives replace a while back right around when this problem started. That message number is that an ANR0877W or ANR8770W? Checked for both those msgno's and didn't come up with any hits in our activity log.

Tape wise we've purchased some additional LTO3 media as needed but didn't see anything odd there, that and just normal ejects and returns from our offsite vendor.

[SleeperGTP]

Did you update your drive serial number when the drive was inserted?

[Mikey D]

Did you update your drive serial number when the drive was inserted?

Drives are all ACSLS and we use Gresham EDT as the external library, outside of ACSLS having to pickup a new serial. That's all that was required from that side when we replaced them.

[picay]

08/17/09 13:12:01 ANR8302E I/O error on drive SL8500 (/dev/rmt/6mt) with volume R01156 (OP=LOCATE, Error Number=201, CC=412, KEY=02, ASC=3A, ASCQ=00, SENSE=70.00.02.00.00.00.00.10.0- 0.00.00.00.3A.00.00.00.94.50.00.00.00.00., Description=M- edia not present in drive). Refer to Appendix D in the 'Messages' manual for recommended action. (SESSION: 14664, PROCESS: 245)

Do you have the meaning of this SENSE code ? and ASC ?
On Gresham, do you have some logged message ?
Clearly, some underlying actor answer something to tsm, which actor ?
So you have to check all the chain, from library and drives to host system, including Gresham
When the problem appear, had you check if media is REALLY missing ?
Did you call TSM ? If so, what is the answer ?