HSZDisk Error Detected

[Shawn Walsh]

I get the following error reported from both servers in a cluster at roughly the same time:

Proliant 7000's
WIN2K SP2
EMA12000

Event Type: Error
Event Source: HS_Service
Event Category: None
Event ID: 1
Date: 1/8/2003
Time: 11:14:32 AM
User: N/A
Computer: TRCRCLUST1
Description:
HSZDisk error detected.
Data:
0000: fd ff ff ff a8 00 00 00 ??????????...
0008: 48 01 00 00 02 00 01 00 H.......
0010: 03 00 00 00 04 00 00 00 ........
0018: 0c 0c 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 33 30 31 30 38 20030108
0038: 31 31 31 34 33 32 2c 37 111432,7
0040: 39 2d 30 35 30 30 00 00 9-0500..
0048: 81 00 29 00 57 69 6e 64 ??.).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 32 00 00 00 00 ck 2....
0078: 21 01 0b 00 54 52 43 52 !...TRCR
0080: 43 4c 55 53 54 31 00 00 CLUST1..
0088: 21 04 10 00 48 53 47 38 !...HSG8
0090: 30 20 20 20 20 20 20 20 0
0098: 20 20 20 20 41 04 05 00 A...
00a0: 56 38 36 50 00 00 00 00 V86P....
00a8: 70 00 06 00 00 00 00 98 p......??
00b0: 00 00 00 00 a1 00 00 00 ....??...
00b8: 00 00 00 00 00 00 00 00 ........
00c0: 00 00 00 00 00 00 00 00 ........
00c8: 01 04 f7 03 41 20 00 00 ..??.A ..
00d0: 00 00 00 00 00 00 00 00 ........
00d8: 00 00 00 00 00 00 20 20 ......
00e0: 20 20 20 20 5a 47 30 31 ZG01
00e8: 38 30 31 36 31 32 56 38 801612V8
00f0: 36 50 08 00 00 00 00 00 6P......
00f8: 00 00 00 00 00 00 00 00 ........
0100: 00 00 00 00 00 00 00 00 ........
0108: 00 00 00 00 00 00 00 00 ........
0110: 06 00 00 00 00 00 00 00 ........
0118: 00 00 00 00 00 00 00 00 ........
0120: 00 00 00 00 00 00 00 00 ........
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 ........


Anyone have a clue as to what is causing it??

TIA!

[JohnWRuffo]

I am not sure if this will help... but take a look.
http://wwss1pro.compaq.com/support/reference_library/viewdocument.asp?countrycode=1000&prodid=52&source=EX020805_CW01.xml&dt=3&docid=13036

[Shawn Walsh]

No, that misses the mark a little, I've got 10K drives in there. Good link though. Still fishing.....

[reg lustenhouwer]

Hi Tia,
HS_Sevice reports "events" that occur on the HSG controller. Some events are serious and some occur under normal operations.

Some examples are:

A raidset that was degraded is now normal (after a rebuild for example)

A fan failed on the controller

A battery that was below normal charge now is fully charged.

A disk has failed in a raid set and now it is in a degraded state.
(These are not real messages but my summarisation of types of events)

I would suggest you look at the visual indecators of the controllers (LEDS of fans ETC)and see if any anomolies and then look at the serial console to see if there is an error.
Use the following
SHOW THIS
SHOW UNIT FULL
SHOW STORAGE FULL
SHOW DISK FULL

If you do not find the issue please tell us under what condition the event occured . Reboot? Power Up after a long power down? Cluster failover ?

I believe that both nodes are just reporting the same event

Reg

[Shawn Walsh]

TIA = Thanks in advance...

Anyway, the controllers show nothing in the way of errors. Yes, both nodes are reporting the same error at the same time which leads me to beilieve there is a problem with the EMA12000 somewhere. The errors are randomly reported throughout the day while the cluster is in normal operating mode (not failing over or being rebooted). The errors usually come in groups of 5 to 7 of the same errors within seconds of each other then it will be fine for up to about 30 minutes. As I said before, there is no performance degradation and I wouldn't even notice the problems had I not pulled the event viewer up by accident.

No error indications on the EMA12000 either, everything is 'green'.
Shawn

[reg lustenhouwer]

Hi Re the "EVENTS" you see you need to make the clear distinction that they are events. Some events are informational ,some events are errors some are a retry or recovery action.

If all visual indecators are green and you see no unusual output from the commands I posted they are most likely informational or recovery events.

You can leave a serial console session conneted to the console serial port and if you see something at the times the "events" occur you may find something more about what is happening.

You can get events on the HSG controller when network errros occur in an Windows cluster which has nothing to do with the storage for example.

Reg

[Shawn Walsh]

These are definately errors, but I can't find anything on the HSG80's. I'll still do more digging. Keep the ideas flowing!!!

[Brad Atchison]

I am also seeing these errors on our system. Our situations seem very similar. We have HSG80 controllers and Proliant 380's running 2000 Advanced Server SP2. I left a console cable on a controller and logged the following error at the same time as the HS_Service errors appeared in the event log:


%EVL--HSG80_TOP>--10-DEC-2002 23:44:44-- Instance Code: 03F20064
Template: 65.(41)
Occurred on 10-DEC-2002 at 23:44:44
Power On Time: 0. Years, 230. Days, 13. Hours, 20. Minutes, 5. Seconds
Controller Model: HSG80
Serial Number: ZG20800693 Hardware Version: E16(2E)
Software Version: V86F-8(BA)
Informational Report
Port: 0. Target: 0.
Instance Code: 03F20064

HSG80_TOP>
%CER--HSG80_TOP>--10-DEC-2002 23:44:44-- SWAP signal cleared - all SWAP-
interrupts re-enabled

No one at Compaq has been able to identify the problem if there even is one. I don't know if you're getting the same errors on your controllers but please keep me informed of your progress.

[Shawn Walsh]

Hmmm, I'll let you know if I encounter that error, but up until now, I haven't seen anything pop up in the term window. How would I go about showing any stored errors or logs???

[Brad Atchison]

These errors were not logged in our controllers or in our SAN Application Server (HSG Element Manager). The only way we saw them was to leave a hyperterm window open and logging.

[Shawn Walsh]

Well, I've had the hyperterm window open for the better part of a month and the only thing that pops up is a screwy error about a fan or battery bad on shelf 5 & 6 then it clears it in the next second. Nothing else at all.

[Joseph Vidal]

It appears that you have a disk error. To find out which one it is, connect to the controller with SN: ZG01801612. When you get a prompt you will need to call up VTDPY. "CTRL-C" will get you the VTDPY-Prompt. "DIS DEV" and return. You will see a table representing your disks. Correspondent to each disk is a table showing various information. The column which we need to pay attention to is the last one, the ?ER? or error table. This is a very good indicator of a defective disk drive. If you see any number other than zero here... Bingo! Mark down which disk has the errors, leave VTDPY with "CTRL-Y" and proceed to replace the disk. (The errors are only logged locally to the disks, which means that if you were to move the drive to another EMA/ESA you would still see them in VTDPY!) Hope this helps you.

[Rob Wunsch]

I have the same problem as well and have not been able to get any answers. In my opionion it is not a disk problem because I have 4 RAID Arrays connected throught the SAN and each server is getting the same error.

[dentistcorn]

Attention! Compaq/HP has a utility that parses only the Application Log of event viewer. I have seen this error many many times ( Event ID 1 HS_Service) and each time this utility has helped me discern the root cause of the problem from failed disks to a myriad of other issues. Other times it would translate to just a warning. I would strongly suggest locating and downloading this utility. It's called Compaq Analyze.

[Shawn Walsh]

Thanks for the Compaq Analyze info... I'm locating now. I think the problem may have been solved: While putzing around behind the racks, I noticed (among all of the other blinking lights) one of the PSU LED's blink occassionally. Sure enough, that power supply was faulty. I strongly recommend having at least one PSU on hand for a quick swap! It looks like the problem has been solved.

Thanks!!!

[Stephen Blanchard]

We are getting the same HSZDisk errors with our HSG80 SAN. HP have replaced 6 disks in our SAN this week. The frequency of the errors has gone down but they still appear when we stress the disks by bulk loading our sql2000 cluster database. I have run the VTDPY program as described above and yes we have disks with numbers in the error column. My question is how do we flush the data in the display so that we can look for new errors. I have done a restart other and restart this on our controllers but the VTDPY error column output has not changed.

TIA

[Stephen Blanchard]

We are getting the same HSZDisk errors with our HSG80 SAN. HP have replaced 6 disks in our SAN this week. The frequency of the errors has gone down but they still appear when we stress the disks by bulk loading our sql2000 cluster database. I have run the VTDPY program as described above and yes we have disks with numbers in the error column. My question is how do we flush the data in the display so that we can look for new errors. I have done a restart other and restart this on our controllers but the VTDPY error column output has not changed.

TIA

[Claude Legare]

Hi!
I'm a new french user so I applogize for my english. I read in the "HSG80 Array Controler V8.7" that : "If the device is swapped or deleted, then, the value clears and resets to 0". But if someone know an other way to clear the error value, I'll be interested because I have somme error to!
Bye.