StoreEver Tape Storage
cancel
Showing results for 
Search instead for 
Did you mean: 

SAN library LTO3 shows NO_HW

SOLVED
Go to solution
Andrew Young_2
Honored Contributor

SAN library LTO3 shows NO_HW

Hi.

I have an MSL5060 tape library with 4 LTO3 drives connected to our SAN via 2 HP StorageWorks e1200-320 4G routers. For the purposes of backup these drives are multipathed to a number of HP-UX servers. The devices show as CLAIMED on HP-UX11.23 servers. However one of the tape drives shows up as NO_HW on the HP-UX 11.31 server. That server is a BL860c with a FC mezzanine card.

I have checked zoning on the SAN switches (if this was a problem then access to the other drive on the router would be a problem as well) and similarly I have checked the configuration on the FC router (e1200) on the tape library itself.

On the server the FC drivers are all B.11.31.0809, the OS version installed is B.11.31.0809 and the Quality Pack is B.11.31.0809.326

Running ioscan -fnC tape brings up the NO_HW error.

There are no problems with the tape drive itself as it is successfully used on the 11.23 servers.

Although I think this is irrelevant to the problem we are using DP6.00 with patches that are less than a month old.

I hope someone can help with this.

Regards

Andrew Y
Si hoc legere scis, nimis eruditionis habes
13 REPLIES
Torsten.
Acclaimed Contributor

Re: SAN library LTO3 shows NO_HW

Hi Andrew,

because of the status of NO_HW the devices were visible before.

So something has changes.

- zoning on switch
- mapping in nsr
- cable connection (unlikely)

Does the switch still see both, the library and the server?

I would start to check the switch and nsr.

Hope this helps!
Regards
Torsten.

__________________________________________________
There are only 10 types of people in the world -
those who understand binary, and those who don't.

__________________________________________________
No support by private messages. Please ask the forum!

If you feel this was helpful please click the KUDOS! thumb below!   
Fabio Ettore
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi Andrew,

even if SAN can seem ok, there is a problem when recognising the tape drive. From the man of ioscan:

NO_HW the hardware at this address is no longer responding

If other HP-UX 11.23 the same drive is correctly recognised then the problem could be from HP-UX HBA to nsr/switch ports.

Check those things (HW cables, ports to which cable are connected, etc.) to solve the problem.

Best regards,
Fabio
WISH? IMPROVEMENT!
Andrew Young_2
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi

Its definitely not a cabling issue - if it were then we'd be able to have other failures along one of the paths but it is unique to this one device.

And its definitely not a zoning issue for the same reason.

As far as checking the NSR mapping are there any suggestions on what to check?

Regards

AY
Si hoc legere scis, nimis eruditionis habes
Torsten.
Acclaimed Contributor

Re: SAN library LTO3 shows NO_HW

You need to compare the WWN of the HBA if it is there in the current mapping (no typos) and if the map is indexed, not auto assigned.

Hope this helps!
Regards
Torsten.

__________________________________________________
There are only 10 types of people in the world -
those who understand binary, and those who don't.

__________________________________________________
No support by private messages. Please ask the forum!

If you feel this was helpful please click the KUDOS! thumb below!   
Andrew Young_2
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi.

I assume you are referring the the mappings on the library and SAN. All the WWNs are correct. That was one of the first things I checked

It is possible with HP-UX 11.31 for the OS to take a tape device off line? I recall reading about this somewhere, but cannot recall where.

Regards

AY
Si hoc legere scis, nimis eruditionis habes
Fabio Ettore
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi,

you're right, it exists scsimgr command to disable a device. Check the condition by

scsimgr get_info -D

and note the line on

"SCSI services internal state"

If it's OFFLINE try to enable it by

scsimgr enable -D

If so next question should be why it was disabled. If this doesn't help though please complete output of

ioscan -fn

Regards,
Fabio
WISH? IMPROVEMENT!
Fabio Ettore
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Maybe the reason why the device could be disabled is in man page of scsimgr:

disable/enable Disables/enables a SCSI object. Upon disable, the
object cannot be used to transfer I/Os. If the
disabled object is a LUN, all pending I/Os are aborted
for that LUN. Currently supported SCSI objects are
LUNs and LUN paths.

The SCSI object can be disabled either administratively
by invoking the disable command, or by the SCSI stack
upon detection of certain types of critical errors.
When this condition has occurred, HP-UX logs a message
in syslog which informs the user that a SCSI object has
been disabled. In the case of the SCSI stack disabling
the SCSI object, the system administrator, after
resolving the problem on the device, can invoke the
enable command to resume data transfers on the SCSI
object.


In that case it should be checked errors/messages in syslog or diagnostic logs.
WISH? IMPROVEMENT!
Andrew Young_2
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi Fabio

Thanks. That was what I was looking for.

Status comes back as UNOPEN.

However when trying to change the status I get the following error:

#scsimgr enable -D /dev/rtape/tape6_BEST
scsimgr: ERROR: enable failed : Function not implemented

Regards

Andrew Y
Si hoc legere scis, nimis eruditionis habes
Fabio Ettore
Honored Contributor
Solution

Re: SAN library LTO3 shows NO_HW

Hi Andrew,

it's the first time I try to enable/disable a tape device and I'm seeing strange things. For example I could disable/enable easily a disk on a my system but get the same error when trying to disable a tape drive (from ESL Tape Library, which Tape Libray do you have?):

# scsimgr disable -D /dev/rtape/tape2_BEST

CRA REPORT SUMMARY:
Critical Resources Analysis(CRA) reported SUCCESS.
No resources usages have been identified during the analysis.

Do you really want to disable (y/[n]) ? y
scsimgr: ERROR: disable failed : Function not implemented

# echo $?
252

which corresponds to

"#define ENOTSUP 252 /* Function not implemented */"

In a way enabling/disabling tape device by scsimgr is not supported but I don't have details. So I'd think your problem is not about tape device disabled.

Have you laredy checked for EMS/SFM events? I suppose you've already checked /var/opt/resmon/log/event.log for events on that drive.

Best regards,
Fabio
WISH? IMPROVEMENT!
Fabio Ettore
Honored Contributor

Re: SAN library LTO3 shows NO_HW

I also found options on ioscan command to check the real status of the lunpath:

ioscan -P health -H

Anyway I'd suggest checking diagnostic log files if they have events for tape drive.

Good luck and let us know if the story gets develops, just for future experience ;)
WISH? IMPROVEMENT!
Andrew Young_2
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi Fabio

The library is an older MSL5060 (one of the old rebadged Tandberg units) that we have upgraded to LTO3.

The disabling did not occur as part of a manual process. I did however find the following entries in the syslog:

Mar 9 11:22:47 hostname vmunix: class : lunpath, instance 10
Mar 9 11:22:47 hostname vmunix: Evpd inquiry page 83h/80h failed or the current page 83h/80h data d
o not match the previous known page 83h/80h data on LUN id 0x0 probed beneath the target path (class
= tgtpath, instance = 6) The lun path is (class = lunpath, instance 10).Run 'scsimgr replace_wwid'
command to validate the change
Mar 9 11:22:47 hostname vmunix:
Mar 9 11:22:47 hostname vmunix: An attempt to probe existing LUN id 0x2000000000000 failed with err
no of 14.
Mar 9 11:22:47 hostname vmunix: class : lunpath, instance 10

From this it looks like when the drive was replaced, the WWID change was not detected or corrected in the configuration.

Everything is now operational again. Thanks for everybody's assistance.

Regards

Andrew Y

Si hoc legere scis, nimis eruditionis habes
Fabio Ettore
Honored Contributor

Re: SAN library LTO3 shows NO_HW

Hi Andrew,

great that the problem is solved, so I understand you issued

scsimgr replace_wwid ...

to solve it. Well done!

Regards,
Fabio

WISH? IMPROVEMENT!
Andrew Young_2
Honored Contributor

Re: SAN library LTO3 shows NO_HW

In summary, HP-UX 11.31 uses WWIDs for all devices and when any SCSI hardware (including tape drives) is changed the WWID will change as well.

Note as well that DP 6 uses the DSF addresses for 11.31 devices and therefor the device ID will change. Update the configuration!



Si hoc legere scis, nimis eruditionis habes