topic Re: HSG80 ERROR PLEASE HELP. in HPE EVA Storage

HSG80 ERROR PLEASE HELP.

krsukuma — Mon, 22 Dec 2008 10:26:04 GMT

Hostname: SWCWDSOPSPROD09

WARNING: Device communication error - ScsiExecCli2:ScsiRecvDiag():ScsiSendDiag() (SP_SCSI: ScsiExecCli)

we get these every day, can anyone please help on this.

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Mon, 22 Dec 2008 14:41:36 GMT

No one to help

Re: HSG80 ERROR PLEASE HELP.

Rob Leadbeater — Mon, 22 Dec 2008 15:03:54 GMT

You might want to provide just a little more information about your environment...

Where are you getting that error ?

Cheers,
Rob

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Mon, 22 Dec 2008 15:39:45 GMT

Hi Rob,

Thanks, This error is an atomatic alert from SWCWDSOPSPROD01.cw.uk.azp>>- SWCWDSOPSPROD01.ukcw.xxxx.net HSG80 11000000000 HSG80 disksdisk41500.1 >NSMTEC>1229525841.1.1>NODE TIME>Dec 17 14:10:40.

This alert is from Host end.Its connected to HSG80. In HSG80 events I cannot find any warnings. No failed disk are showing.Host are online.

Unix team told me this certainly in some days a drive is going to be failed, so this might be a early warning, but I am not sure abt that.

Re: HSG80 ERROR PLEASE HELP.

Rob Leadbeater — Mon, 22 Dec 2008 16:05:20 GMT

Again, that tells us little about the environment...

What is the hardware configuration ?

A "SHOW DISK41500" from the HSG80 might be useful to see.

Cheers,
Rob

Re: HSG80 ERROR PLEASE HELP.

Rob Buxton — Tue, 23 Dec 2008 00:01:38 GMT

I'm guessing this error is from SWCC. I've seen something similar from SWCC running on our old VMS Servers monitoring HSD Controllers.

In our case it's correctly reported problems where the controller has gone offline.

Again, just guessing, in this case it's either the disk or one of the paths to the disk that's causing problems. The monitor on the server is detecting that it can no longer communicate with the device.

More detail on how the HSGs are connected to the server might help. Are there multiple paths etc.

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Tue, 23 Dec 2008 09:11:19 GMT

show disk output:
DISK41500 disk 4 15 0 SPARESET
COMPAQ BD03663622 BDC4
Switches:
NOTRANSPORTABLE
TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)
Size: 71114623 blocks

Show this:
Controller:
HSG80 ZG02601364 Software V87F-7, Hardware E12
NODE_ID = 5000-1FE1-000B-4750
ALLOCATION_CLASS = 0
SCSI_VERSION = SCSI-2
Configured for MULTIBUS_FAILOVER with ZG10706996
In dual-redundant configuration
Device Port SCSI address 7
Time: 23-DEC-2008 09:59:26
Command Console LUN is lun 0 (NOIDENTIFIER)
Host Connection Table is NOT locked
Smart Error Eject Disabled
Host PORT_1:
Reported PORT_ID = 5000-1FE1-000B-4753
PORT_1_TOPOLOGY = FABRIC (fabric up)
Address = 010A00
Host PORT_2:
Reported PORT_ID = 5000-1FE1-000B-4754
PORT_2_TOPOLOGY = OFFLINE (offline)
NOREMOTE_COPY
Cache:
128 megabyte write cache, version 0022
Cache is GOOD
Unflushed data in cache
CACHE_FLUSH_TIMER = DEFAULT (10 seconds)
Mirrored Cache:
128 megabyte write cache, version 0022
Cache is GOOD
Unflushed data in cache
Battery:
NOUPS
FULLY CHARGED
Expires: 18-APR-2009

Re: HSG80 ERROR PLEASE HELP.

Rob Leadbeater — Tue, 23 Dec 2008 12:53:29 GMT

Hi,

> Reported PORT_ID = 5000-1FE1-000B-4754
> PORT_2_TOPOLOGY = OFFLINE (offline)

This looks like it *might* be the problem...

I'll guess that you might have a SAN switch down, but again as you've failed to give us any details on the hardware configuration, that's complete speculation...

Cheers,

Rob

Re: HSG80 ERROR PLEASE HELP.

Uwe Zessin — Tue, 23 Dec 2008 13:00:43 GMT

No, it means that the port has never been configured or manually been disabled:
CLI> SET THIS_CONTROLLER PORT_2_TOPOLOGY = OFFLINE

A disconnected link looks like this:
PORT_1_TOPOLOGY = FABRIC (connection down)

Re: HSG80 ERROR PLEASE HELP.

marsh_1 — Tue, 23 Dec 2008 13:35:58 GMT

hi,

as for the disk , at least it's a spare disk. has it any error lights on the drive ?
if you 'run fmu' on the hsg80 what does the last error say 'show last most' ?

Re: HSG80 ERROR PLEASE HELP.

marsh_1 — Tue, 23 Dec 2008 15:05:40 GMT

hi,

also what does 'show fail' say on the hsg80 ?

just thinking of spareset smaller than failed drive on array ?

good luck

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Wed, 24 Dec 2008 11:56:21 GMT

Name Storageset Uses Used by
------------------------------------------------------------------------------

FAILEDSET failedset
Switches:
NOAUTOSPARE

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Wed, 24 Dec 2008 12:01:26 GMT

On HSG80 I dont see any failures, and more in controller host ports two is offline. I hav e5 more HSG with this kind, I dont get any alerts for this, and I beleive one host is enough, as second host port is kept for reduncy.

Re: HSG80 ERROR PLEASE HELP.

marsh_1 — Wed, 24 Dec 2008 14:42:30 GMT

HI,

if you have no errors in fmu and no fault lights on the drive, unless you want to run dilx against the unit which would impact performance on the hsg80 or initialize it and create a lun then present it to the host,write to it etc then delete it and test it that way - the issue would seem to be with the swcc implementation on the host , maybe reinstall or upgrade the version.

fwiw

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Wed, 24 Dec 2008 15:29:33 GMT

I did not run the FMU command, I tried with that command it gives me a warning, I have attached the snapshot, please tell me 100% sure if I run the command will it create any problem in HSG80.

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Wed, 24 Dec 2008 15:31:52 GMT

One more thing in thie HSG I dont create any Luns,I am Just montiering for hardware failures, so i will not able to create any luns in this.

Re: HSG80 ERROR PLEASE HELP.

marsh_1 — Wed, 24 Dec 2008 18:48:39 GMT

hi,

FMU is the fault management utility , i normally run it from the cli on the hsg80, it won't do any harm to your hsg80.
it's one of a number of utilities available on the hs/z series controllers such as frutil, vtdpy and hsutil for performance monitoring, replacement batteries , upgrading firmware etc.
i take it you are doing this through swcc ?

hope this helps

Re: HSG80 ERROR PLEASE HELP.

Uwe Zessin — Wed, 24 Dec 2008 19:14:36 GMT

> ScsiExecCli

This looks like a communications error between the SWCC agent and the HSG80 controller.

I suggest to check the Windows event log and the server's switch port counters for errors.

Re: HSG80 ERROR PLEASE HELP.

krsukuma — Tue, 30 Dec 2008 11:17:11 GMT

Hi mark hunt,

I run the FMU commands, SHOW LAST ALL,SHOW LAST MOST and SHOW MEM ALL,
I have attached the doc, can you please help whether the HSG is reporting any logs on this.

Re: HSG80 ERROR PLEASE HELP.

Rob Leadbeater — Tue, 30 Dec 2008 17:41:32 GMT

Can you supply the same FMU logs, and "show this" for the other controller....

Cheers,
Rob