Riddle me this ... tape drive NO_HW

[Todd McDaniel_1]

Okay, I have 19 tape drives as part of a tape san for netbackup... visible by 5 "media servers" which backup 17 boxes.

My problem is that the drive shows NO_HW but has been available in the past. Our Storage management folks, who manage the Tape san btw, say that it has been writing so that precludes it from being on their end... STK came out and saw that the tape device was writing at that point.

MY point is they, Storage Management, keep asking me to "rediscover the drive"... which I have numerous times wiht insf and ioscan.

My dilemma is they say its the OS and I say it cannot be the OS...

I dont know what to try to get this drive back online... They say they have done everyting they can do. ditto for me.


These are remote drives btw... Here is what they look like.

# ioscan -fnC tape |grep STK
tape 17 0/5/0/0.101.18.255.13.12.0 stape CLAIMED DEVICE STK 9840
tape 18 0/5/0/0.101.19.255.13.10.0 stape CLAIMED DEVICE STK 9840
tape 19 0/5/0/0.101.20.255.13.9.0 stape CLAIMED DEVICE STK 9840
tape 7 0/5/0/0.101.21.255.13.6.0 stape CLAIMED DEVICE STK 9840
tape 1 0/5/0/0.117.22.255.14.1.0 stape CLAIMED DEVICE STK 9840
tape 2 0/5/0/0.117.23.255.14.8.0 stape CLAIMED DEVICE STK 9840
tape 3 0/5/0/0.117.24.255.14.15.0 stape CLAIMED DEVICE STK 9840
tape 4 0/5/0/0.117.25.255.13.6.0 stape CLAIMED DEVICE STK 9840
tape 5 0/5/0/0.117.26.255.13.9.0 stape CLAIMED DEVICE STK 9840
tape 6 0/5/0/0.117.27.255.14.0.0 stape CLAIMED DEVICE STK 9840
tape 20 1/8/0/0.101.22.255.13.5.0 stape CLAIMED DEVICE STK 9840
tape 15 1/8/0/0.101.23.255.13.9.0 stape CLAIMED DEVICE STK 9840
tape 16 1/8/0/0.101.24.255.14.0.0 stape CLAIMED DEVICE STK 9840
tape 9 1/8/0/0.118.22.255.14.1.0 stape CLAIMED DEVICE STK 9840
tape 10 1/8/0/0.118.23.255.14.2.0 stape CLAIMED DEVICE STK 9840
tape 11 1/8/0/0.118.24.255.14.4.0 stape CLAIMED DEVICE STK 9840
tape 12 1/8/0/0.118.25.255.14.8.0 stape NO_HW DEVICE STK 9840
tape 13 1/8/0/0.118.26.255.14.15.0 stape CLAIMED DEVICE STK 9840
tape 14 1/8/0/0.118.27.255.13.6.0 stape CLAIMED DEVICE STK 9840

[RAC_1]

rmsf "device files of the drive that is not visible form OS"

ioscan -u -Ctape

insf -ev -Ctape

Anil

[Todd McDaniel_1]

How exaclty would I remove all these devcies?

I know how to use the rmsf command but do I remove all these?


tape 12 1/8/0/0.118.25.255.14.8.0 stape NO_HW DEVICE STK
9840
/dev/rmt/12m ........ /dev/rmt/c51t8d0BESTn
/dev/rmt/12mb ...... .. /dev/rmt/c51t8d0BESTnb
/dev/rmt/12mn ...... .. /dev/rmt/c51t8d0D3480
/dev/rmt/12mnb ...... .. /dev/rmt/c51t8d0D3480b
/dev/rmt/c51t8d0BEST ...... /dev/rmt/c51t8d0D3480n
/dev/rmt/c51t8d0BESTb ...... /dev/rmt/c51t8d0D3480nb

[RAC_1]

Hold on for a minutes, before you delete those files with rmsf.

Can you do
mt /dev/rmt/12m stat

does it give any error??

[Todd McDaniel_1]

# mt -f /dev/rmt/12mn stat
on tape open: No such device or address

[RAC_1]

OK,
Check the connections - loose, not properly fitted etc. IF everything is OK, then

rmsf /dev/rmt/12m

ioscan -u -Ctape

insf -ev -Ctape

Post

Anil

[Todd McDaniel_1]

No avail bro...

I still have no hw on them and I dont have physical access to these drives... They are in ST Louis and I am in Dallas.

My Storage management has looked at the drives with STK, I would HOPE they checked the cabling...

[RAC_1]

Does, ioscan -u reports that drive??

Anil

[Todd McDaniel_1]

yes it does...

tape 12 1/8/0/0.118.25.255.14.8.0 stape NO_HW DEVICE STK 9840


# ioscan -u -Ctape
H/W Path Class Description
===============================================================
0/0/1/0.0.0 tape HP C1537A
0/5/0/0.101.18.255.13.12.0 tape STK 9840
0/5/0/0.101.19.255.13.10.0 tape STK 9840
0/5/0/0.101.20.255.13.9.0 tape STK 9840
0/5/0/0.101.21.255.13.6.0 tape STK 9840
0/5/0/0.117.22.255.14.1.0 tape STK 9840

-------------------------------------------
0/5/0/0.117.23.255.14.8.0 tape STK 9840
-------------------------------------------

0/5/0/0.117.24.255.14.15.0 tape STK 9840
0/5/0/0.117.25.255.13.6.0 tape STK 9840
0/5/0/0.117.26.255.13.9.0 tape STK 9840
0/5/0/0.117.27.255.14.0.0 tape STK 9840
1/8/0/0.101.22.255.13.5.0 tape STK 9840
1/8/0/0.101.23.255.13.9.0 tape STK 9840
1/8/0/0.101.24.255.14.0.0 tape STK 9840
1/8/0/0.118.22.255.14.1.0 tape STK 9840
1/8/0/0.118.23.255.14.2.0 tape STK 9840
1/8/0/0.118.24.255.14.4.0 tape STK 9840
1/8/0/0.118.25.255.14.8.0 tape STK 9840
1/8/0/0.118.26.255.14.15.0 tape STK 9840
1/8/0/0.118.27.255.13.6.0 tape STK 9840

[RAC_1]

As you told the drive may be usable, in use, but as this is SAN drive, other hosts might be able to see it. But seems that from this particular box, you won't see it. Could be connection problem from your box to drive.

Anil

[Todd McDaniel_1]

Right now, it is not visible from any of the boxes that should see it.

I can only take STK at their word that they saw the tape writing as I have no direct access to it.

They say it has been visible at some time but I seriously doubt that.

[John Poff]

Hi,

Another thing to check is to compare the firmware version on the problem drive against one or two of the drives that are working. Maybe they replaced a bad drive with one that has new or old firmware which isn't playing well with HP.

JP

[Michael Tully]

If this drive is part of a library as such, and it forms with others in a bridge, you should see it unless there is something wrong with the bridge. But if it is working from other sources, this cannot be it. Have you looked at zoning? Usually when you see drives and then not see them but they still physically work, then perhaps a zone has been changed? ....

[Dave La Mar]

Todd -
One more thought along John's line. If any of the intermediate connect equipment was replaced is that piece current on firmware or was it rebooted to acknowledge the drive?
We have seen similar on fibre connected drives where the gbit was changed and had to be rescanned by a reset of the bridge.

This is a good one. I'll ejoy following theis thread to fruition.
BTW are drives are not remote but have seen similar related to firmware and reboot.

Best regards,

dl

[Todd McDaniel_1]

We are having a meeting today today with Storage folks...

I did run the suggested commands after RMSF command and no avail... still the same problem.

Mike,

I will let you know what they say. I doubt the zoning has changed since I can see the other drives... but no telling what happened when they replaced the drive.


Dave,

I will ask about any other changes, but tend to doubt it as well...



I am really stumped on this one but one thing I AM sure about it this aint on my OS, must be on the other end somewhere.