Fibre Channel not talking.

[UI HP Sysadmins]

Configuration:

K400/2 cold install HP-UX 11i
A3404A <---> A5236A

NO Patches (yet)
I had 11.00 fully patches still had the same problem.


Problem:
I can see the disk, I just can not make them do anything. What am I missing?

expample: dd if=/dev/rdsk/c9t12d0 of=/dev/null bs=1024
NEVER RETURNS.

pvcreate
NEVER RETURNS.

[harry d brown jr]

UI,

What kind of disk subsystem do you have? Did you use the Dec2001 or the latest Mar2002 install media? Also, you should apply all hardware and OS patches.

Also, on your "dd", try this:

dd if=/dev/rdsk/c9t12d0 bs=1024 count=5 | od -bc

live free or die
harry

[Jeff Schussele]

What does diskinfo return?

I've had "seen" disks that couldn't respond until reseated in their enclosure.

HTH,
Jeff

[UI HP Sysadmins]

# diskinfo /dev/rdsk/c9t12d0
SCSI describe of /dev/rdsk/c9t12d0:
vendor: SEAGATE
product id: ST118202FC
type: direct access
size: 35328458 Kbytes
bytes per sector: 520

# diskinfo /dev/rdsk/c9t13d0
SCSI describe of /dev/rdsk/c9t13d0:
vendor: SEAGATE
product id: ST136403FC
type: direct access
size: 69843429 Kbytes
bytes per sector: 520


I have reseated the disk. Still the same.

[UI HP Sysadmins]

The Disk Subsystem is a FC10.

The install media was Sep2001.

I am going after the patches now.

I have (4) FC10 but only one is hooked to the system currently. How ever in the past I have gotten the same result(prblem) from all of them.

[Jeff Schussele]

Something doesn't seem right here.

Aren't those 18 & 36 GB drives & look at the capacity they're reporting.

Am I wrong here?

Jeff

[Helen French]

Hi,

1) R u able to see the disks from SAM ? Did you tried doing something through SAM ?

2) what is the status of the disk when looking from STM ?

3) Did you try applying the latest patches ?

Some thoughts ...

Shiju

[UI HP Sysadmins]

The capacity of the disks is incorrect you are correct. I think that may be part of the problem. I am unable to get it to report correctly yet.

I will be patching soon.

Sam sees the disk and the controller fine.

Sam also report the disk size correctly.

In STM:
Everything is Green (ok) if I just get information, but if I ask to "Exercise" the disk it hangs. The system is still functional but the process never returns.

[harry d brown jr]

UI,

jeff hit it on the head, your system is seeing the disks as something they are not. How many 18gb and/or 36gb drives do you have? Also, on the back of the FC10, there are TWO (2) dials, one at the top middle and one at the bottom middle, what are they set to? Also, there are FOUR Fibre Channel (FC) ports. Two are Primary, and two are Alternate. I don't remember exactly where they are, but they are labeled, and I think I rmember the top left (from the rear) is a primary, and the bottom right is another primary. What "port" do you have your FC cable plugged into (it actually doesn't matter - the following explains)

I have TWO FC10's, which are not in any kind of loop nor are they on a switch. I have an N-class server with 6 virtual partitions, where three partitions are using the same FC10. two are using the primary ports, and the thrid is using the alternate port. I just have to make sure I don't accidentally try to assign a disk being used by one partition to another.

live free or die
harry

[Jeff Schussele]

BTW - kudos to UI for attaching relevent info on org msg.

Boo on us for not noticing - sorry.

Harry - his ioscan shows just the 2 disks. His fcmsutil on fcsm0 shows IN_LOOP. Is this what happens if he has the FC10 hooked to both primaries?

Jeff

[harry d brown jr]

jeff,

If it's really a FC10 without a switch, then it doesn't matter if you plug into the primary ports or the alternate ports. Read my last post about what I did with my N-class.

My device is /dev/td1, because it's a pci tachyon fc adapter, and it shows PRIVATE_LOOP.

live free or die
harry

[UI HP Sysadmins]

I have
(17) 36 GB disk
(8) 18 GB disk
(15) 9 GB disk
but I only have two disk in the FC10 currently to make the problem simpler.
Both dials on the back of the FC10 are set to "1". I only have the system attatched to one of the FC10 to make it easier to diagnose also.
The system is connected to the buttom card in the FC10 via the Primary port, only one connection currently. You are correct on the port location, it is lower Right.

I have been told that the Fibre Channel devices self terminate, so as long as I do not put a GBIC in the expansion port the loop is closed - is this correct?

Thank you.

[UI HP Sysadmins]

I have finally got the system patched and did some more tests. I still get the same results. Does anyone have any more ideas. See attachment for list of patches.

info from STM:

Hardware path: 8/0.8.0.255.0.12.0


Product Id: ST118202FC Vendor: SEAGATE
Device Type: SCSI Disk Firmware Rev: HP03
Device Qualifier: SEAGATEST118202FC Logical Unit: 0
Serial Number: LK34679800002912K1JP
Capacity (M Byte): 17519.76
Block Size: 520
Max Block Address: 35328457
Error Logs
Read Errors: 0 Buffer Overruns: N/A
Read Reverse Errors: N/A Buffer Underruns: N/A
Write Errors: 0 Non-Medium Errors: 0
Verify Errors: 0


# mediainit -v /dev/rdsk/c9t12d0
mediainit: initialization process starting
mediainit: locking SCSI device
mediainit: initializing media
mediainit: initialize media command failed - I/O error

[Steven Sim Kok Leong]

Hi,

Can't be of much help here. But, have you tried troubleshooting by elimination?

1) Change the fibre cores

If you have only one disk attached to each fibre pair, then is there a possibility that tx works but rx does not?

2) Change the disk.

3) Add inline termination if possible.

4) Connect this disk array to another system.

Check if any of the above works.

Hope this helps. Regards.

Steven Sim Kok Leong

[UI HP Sysadmins]

I think we have figure out what the problem is. It shows up on the last line of the diskinfo.

# diskinfo /dev/rdsk/c9t13d0
SCSI describe of /dev/rdsk/c9t13d0:
vendor: SEAGATE
product id: ST136403FC
type: direct access
size: 69843429 Kbytes
bytes per sector: 520 <<<--- Shoud be 512

These disks are formatted for a virual array. I am using them directly connected.

Now that we know what the problem is, does any one know the solution: How do I reformat these disks?

mediainit does not seem to work.
# mediainit -v /dev/rdsk/c9t12d0
mediainit: initialization process starting
mediainit: locking SCSI device
mediainit: initializing media
mediainit: initialize media command failed - I/O error

Any ideas.

[Jeff Schussele]

Try the format command.

format /dev/rdsk/c9t12d0

I sounds like it can perform sector size change - see the following link:

http://docs.hp.com/cgi-bin/fsearch/framedisplay?top=/hpux/onlinedocs/B9106-90008/B9106-90008_top.html&con=/hpux/onlinedocs/B9106-90008/00/01/119-con.html&toc=/hpux/onlinedocs/B9106-90008/00/01/119-toc.html&searchterms=format&queryid=20020318-095941

HTH,
Jeff

[UI HP Sysadmins]

Final posting.

Problem solved.

I was able to get a program to reformat the drives to 512 Kbytes per sector from HP. I am formatting drives right now.
I would like to thank every one that helped me with this problem.

THANK YOU!!!

-- Ray