Changed Tape Drive on Library w/o Dynamic WWN

[mtfahey]

Hello,
We had a drive fail on our StorageTek Lib. We don't have Dynamic WWN's. Our LAN guys rezoned the tape drive, and also removed the old wwn associated with the FCID and assigned the the new wwn to the old FCID. Now I have NO_HW for this device after an ioscan -fn. This happening in HP-UX 11.11 and 11.23, however, I've 2 servers that see the tapes fine-they have been rebooted since the change - POINTS WILL BE AWARDED

[Steven E. Protter · Accepted solution]

Shalom

The tape drive appears as a different device because you rezoned.

The better thing to do would be to change the WWN in the configuration to the new one.

ioscan -fnC tape

You will see the new device.

The old one can be removed with rmsf

SEP

[mtfahey]

I have done both the ioscan -fnC tape and rmsf -H - Our goal is to not change the way the device is seen(the original device path) - we use the device in Data Protector, and that requires us to change the settings there as well. I initially rmsf'd all the devices(3 devs-this is the second time this drive has failed)then tried to ioscan -fnC tape - I also tried to use insf -e and insf -H

[Torsten.]

So the new drive has a new WWN now?

Have a look into the syslog file!

You will see some entries suggesting you to run

fcmsutil /dev/... replace_dsk ...

to allow the new device to be accepted by the driver.

example:

fcmsutil /dev/td0 replace_dsk 0x00ad

run ioscan again.


A reboot will give the same result...

[mtfahey]

I have seen the those messages in syslog - these are production boxes for 2 of the servers that run 11.11 - but not all the servers give that message. Ex:

Apr 7 13:35:57 sdc9011 vmunix: 1/2/0/0: Unable to access previously accessed device at nport ID 0xc5200.
Apr 7 13:39:58 sdc9011 vmunix: 1/2/0/0: 'World-wide name' (unique identifier) for device at nport ID 0xc5200 has
Apr 7 13:39:58 sdc9011 vmunix: changed. If the device has been replaced intentionally, please use the
Apr 7 13:39:58 sdc9011 vmunix: fcmsutil(1M) command's replace_dsk option to allow the new device to be used.
Apr 7 13:39:58 sdc9011 vmunix: 1/2/0/0: Fibre Channel Driver detected a parse error in the FLOGI/PLOGI response
Apr 7 13:39:58 sdc9011 vmunix: returned by nport ID 0xc5200. FLOGI/PLOGI Fail Code = 0x6.
But as I say I don't receive this in all the servers, just 3, that I've seen.

I do see the drives correctly on 2 servers that were rebooted after the drive was changed - But rebooting all the servers is really not an option as you can imagine.

[Torsten.]

You don't need to reboot.

Just get the device of the fiber HBAs from ioscan -fn and run the command as suggested.

[mtfahey]

Torsten - let me ask you this - on servers that I am not getting this message should I run the the command with the 'replace_dsk' option?

[Torsten.]

From the man page (fcmsutil):

"This option is used to specify that no authentication should be performed the next time we communicate with the device. This option is to be used by system administrators for replacing a disk, with another (with the same N_Port_ID). "


Background: the driver will NOT allow to access the device if the nPort ID stay the same and the WWN will change at the same time. With this command you tell the driver not to check this and allow the device to be accessed.

If you reboot, you don't need to run this command, because the driver will not "remember" the old device.

[Aashique]

Hi Mtfahey,
For you situation i faced last 2 days ago.My 4 drive was showing NO_HW. After that i rebooted the tape library.
After rebooted my problem resolved.

Thanks & Regards

Aashique

[Torsten.]

Rebooting the library will NOT change anything if the new drive has a new WWN. Don't do this. I guess your library is not a small one ...

[mtfahey]

I have a StorageTek L1400, when we put the new in, our LAN guys zoned the new drive in, then mapped the original fcid: 0x0c5200, to the new device - the new device has its own wwn but has the old fcid.

[Torsten.]

So you should be able now to solve your "problem".

[mtfahey]

Well, that is how it has been since the beginning, we want to be able to use the old fcid so that when we have to change the drive out it will be there with the old hardware path - as I mentioned before it was successful on two servers that were rebooted, but not on any of the others

[Torsten.]

Well, did you already run the "fcmsutil" command for all the fc HBAs???

[mtfahey]

Apr 2 16:18:21 sdc9014 vmunix: 1/8/0/0: Unable to access previously accessed device at nport ID 0xc5200

fc 1 1/8/0/0 td CLAIMED INTERFACE HP Tachyon XL2 Fibre Channel Mass Storage Adapter
/dev/td1


fcmsutil /dev/td1 replace_dsk 0xc5200
Disk at nportid 0x0c5200(Loop_id 126) will not be authenticated

I did not have success with the command on this test server - have I done the commands correctly?

[Torsten.]

What is the status for the tape in ioscan after running the command? It should be CLAIMED again. Watch the syslog file.