HSZDisk error detected

[Lawrence Bowen]

Hi all, one of my SAN attached servers has started to generate this error in the app. log.

I have copied the values below from the event, does anyone know how to decipher this?

thanks - Lawrence Bowen

0000: fd ff ff ff ac 00 00 00 ýÿÿÿ¬...
0008: 4c 01 00 00 02 00 01 00 L.......
0010: 03 00 00 00 04 00 00 00 ........
0018: 07 00 00 00 00 00 00 00 ........
0020: 00 00 00 00 00 00 00 00 ........
0028: 05 00 00 00 41 00 18 00 ....A...
0030: 32 30 30 34 30 33 31 39 20040319
0038: 30 37 30 33 33 38 2c 35 070338,5
0040: 37 2d 30 30 30 30 00 00 7-0000..
0048: 81 00 29 00 57 69 6e 64 Â .).Wind
0050: 6f 77 73 20 4e 54 20 35 ows NT 5
0058: 2e 30 20 62 75 69 6c 64 .0 build
0060: 20 32 31 39 35 20 53 65 2195 Se
0068: 72 76 69 63 65 20 50 61 rvice Pa
0070: 63 6b 20 34 00 00 00 00 ck 4....
0078: 21 01 0d 00 52 53 4e 57 !...RSNW
0080: 2d 4d 52 48 2d 48 30 31 -MRH-H01
0088: 00 00 00 00 21 04 10 00 ....!...
0090: 30 00 30 00 31 00 46 00 0.0.1.F.
0098: 45 00 31 00 30 00 30 00 E.1.0.0.
00a0: 41 04 05 00 30 00 38 00 A...0.8.
00a8: 00 00 00 00 f0 00 06 06 ....ð...
00b0: 3b 35 35 98 00 00 00 00 ;55Â ....
00b8: 17 03 d2 80 00 26 00 00 ..Ã Â .&..
00c0: 01 00 17 03 00 00 06 0e ........
00c8: 00 00 00 00 0a 45 28 03 .....E(.
00d0: 51 20 00 28 80 00 00 00 Q .(Â ...
00d8: 00 00 00 00 00 00 00 00 ........
00e0: 00 00 20 20 20 20 20 20 ..
00e8: 5a 47 30 32 37 30 30 36 ZG027006
00f0: 33 38 56 38 36 46 01 00 38V86F..
00f8: 00 00 00 48 50 42 36 42 ...HPB6B
0100: 46 30 37 32 38 35 41 33 F07285A3
0108: 36 20 20 20 20 20 20 00 6 .
0110: 00 00 00 00 f0 00 01 06 ....ð...
0118: 3b 35 35 0a 00 00 00 00 ;55.....
0120: 17 03 d2 80 00 26 00 00 ..Ã Â .&..
0128: 00 00 00 00 00 00 00 00 ........
0130: 00 00 00 00 00 00 00 00 ........
0138: 00 00 00 00 00 00 00 00 ........
0140: 00 00 00 00 00 00 00 00 ........
0148: 00 00

[Uwe Zessin]

Looks like a message about a disk drive: firmware 'HPB6', modell 'BF07285A3(6)'. If this message is still coming up I suggest you connect a terminal or PC with terminal emulator to your HSG80 with serial number 'ZG02700638' and collect the output. It is much easier to read.

[Lawrence Bowen]

Hi thanks for the reply, however what output exactly should i get from the HSGs?

thanks

Lawrence

[Darcy_7]

If you log a call with the storage team and send them the application log in evt format, they can decode the event log and translate the hs_service errors to something more meaningfull.

[Uwe Zessin]

Lawrence,
it is a text output, but the exact message depends on the type of event. Feel free to post it (as an attachment, please) and I am sure somebody of us can help further.

Apart from that I agree with Darcy.

[Michael Schulte zur Sur]

Hi,

if you keep a notebook attached to the console port of the controller, you would be able to catch any messages, the controller generates and therewith what causes the events on your windows machine.
Sometimes the controller gives you a short message along with the prompt.

greetings,

Michael

[Uwe Zessin]

Good idea. Let me point out the messages are not duplicated between controllers, so you need to connect to both if you want to catch all messages.

[Lawrence Bowen]

Thanks for the info people. Do the controllers not store a log of these errors? Or can they at least be configured to store a log?

I would attach the .evt log but it exceeds the 1MB limit.... :)

-Lawrence

[Michael Schulte zur Sur]

Hi,

where should it do so? It would have to send it by snmp or so. Can you cut a sample and post it as attachment. I assume, that the messages are reoccurring events.

greetings,

Michael

[Uwe Zessin]

As far as I know there are is only a limited number of controller errors stored internally, but not for disks.

[Lawrence Bowen]

Ok, I can't shorten the length of the .evt file without clearing the log, and i am not allowed to do that. I've tried filtering the view and then saving, but it still creates a full-size file. Can I ask why the info provided in my original message is not adequate? What additional info will the .evt file contact that the event doesn't?

-Lawrence

[Michael Schulte zur Sur]

Hi,

one difference is that windows does not know anything about the internas of the hsg.
I assume, that you have saved the log to disk. So you could copy and paste different sections, add them to a file and post it as attachment.

Michael

[Mike Naime]

I'm not sure exactly how much memory is related to the FMU logs, but my TAM normally asks for a dump of

SHOW THIS
SHOW OTHER
RUN FMU
SHOW TIME
SHOW LAST MOST
SHOW LAST ALL
SHOW MEM ALL
EXIT

[Lawrence Bowen]

Thanks for the advise chaps, however....

As suggested, I connected a serial lead to both controllers and ran 2 hyperterm sessions to capture any text from the HSGs. This morning there are a further 4 "HSZDisk error detected" events on the server, but not a single message from either of the two controllers....

So, more assistance is required please! I have much more faith in this forum than HP support since every call I have ever logged with them has just been fobbed off in some way, and eventually solved by either myself or someone else's suggestions from this forum :)

thanks

-Lawrence Bowen

[Michael Schulte zur Sur]

Hi,

usually there is a readable part in the event in windows. What does that say?
Is the event in windows classified as error?

greetings,

Michael

[Lawrence Bowen]

Hi,

The description says only:
"HSZDisk error detected."

The rest is copied into the 1st message of this thread.

thanks for your interest.

-Lawrence Bowen

[Lawrence Bowen]

Also, it is not possible to shorten the .evt file. Have you ever looked at an .evt file? It is not text! (some of it is, but there is a lot of code too)

thanks

-Lawrence Bowen

[Mike McCreave]

Hi.

I had this same problem but the Data section was slightly different than yours. However I did some digging arround and found that my HSG80's battery's were coming to the end of life and needed replacing. Couldnt logically tie this in with the errors so I downloaded the Compaq Analyser tool - Compaq being Compaq make it incredibly good fun and make you hunt it down but the download site for it is: http://h18000.www1.hp.com/support/svctools/webes/webesDownload.asp

This is the Webes package (there's documentation and everything but its very easy to figure out. After down load - Install it on your PC. This takes some time as it asks you company, seriel number and other info along the way.

Once its installed go to the effected server and save your event log file (anyname.evt) copy this log file to C:\Program Files\hp\svctools.

Then navigate to Startmenu - Programs - Hewlet Packard Service Tools - System Event Analyser - 'System Event Analyser'

The Analyser will start up, enter your 'login' when prompted and then select 'Other Logs at the bottom of the left hand pane.

In the Left hand pane you'll see Svctools and your anyname.evt log file will be in there - Select it and click ADD at the bottom of the screen. this will apear under 'Other Logs' on the left hand side - Click on the 'anyname.evt' file in the left pane and it will scan it and display what the errors mean on the right - click on one of the links to get full descriptions etc of the fault.

Mine did turn out to be the Cache battery that was failing and this gave me the part numbers and fault codes to call Compaq with and get them replaced.

Good Luck!

[Craig_83]

Hi,

Have you enabled verbose logging on the controllers?

You can enable it by doing the following on each controller:

run fmu
FMU> set event_logging verbose
FMU> set last_failure_logging verbose
FMU> exit

[Fab_3]

I have a RA8000 attached to Proliant 5500 with NT4 Sp6. Every 24 hours in event viewer i receive four messages: "hszdisk error detected". After 1 minute i receive the same four messages. Next day same behavior and so on.
Any ideas ?

[Fab_3]

i followed Mike's hint (System event analizer). My problem is battery near end of life.

[Dorothy Tartt]

I received the same error and we have an HSG80 SAN, as well. After running the analyzer, it came up with some older information that was generated in November, even though I had event log errors for February 10th, 11th, 12th, and 13th. It was kind of weird, but when I went into the Hyperterminal to run a SHOW THIS_CONTROLLER command and SHOW OTHER_CONTROLLER it immediately stated that the cache battery was near the end of its life and needed to be replaced. Called HP and verified that this was the problem. The older errors were picked up from November for some odd reason, but they were valid errors, which were decoded as being related to a Power supply issue. In November we had a Power Outage so that would have been correct. The System Analyzer tool is very handy. Thanks!