dl360 g5 p400i SAS mirror "inconsistent stripe"

[Kapp Kapter]

I am working with HP's ISS Solution center on this one as well.

Install this server last week and am up-to-date with latest drivers/firmware(8.2 cdrom)/service packs as of Monday, Aug 25.

I am receiving an "inconsistent stripe"
from Cissesrv (see below). 2 mirrored drives (146gig 10k SAS)

Anyone else see this? know how to resolve it? I sent an offline adu-report.zip to HP, which they said was AOK (other than the missing battery but one does not come with this model...we are only using Read only cache)



Event Type: Warning
Event Source: Cissesrv
Event Category: None
Event ID: 24607
Date: 8/28/2008
Time: 5:21:36 AM
User: N/A
Computer: JRWMAIL
Description:
The event information received from array controller [Embedded] was of an unknown or unrecognized class.

An excerpt of the controller message is as follows: Inconsistent stripe, LDrv=0 LBA=0x00081B100-0x00081B1FF.

[KarloChacon]

hi

just for test purposes what if downgrade driver or firmware for controller and check if you got the same issue since you have the latest

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=1121486&prodNameId=3288144&swEnvOID=1005&swLang=8&taskId=135&swItem=MTX-cfa764aea73141c9892f909764&mode=3

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=1121486&prodNameId=3288144&swEnvOID=1005&swLang=8&taskId=135&swItem=MTX-d7cf9e5b20e8423983267549a3&mode=3

regards

[KarloChacon]

hi again

even downgrade this

HP ProLiant Smart Array SAS/SATA Event Notification Service for Windows Server 2003/2008

since it's the source -- CISSESRV --


http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=1121486&prodNameId=3288144&swEnvOID=1005&swLang=8&taskId=135&swItem=MTX-9019921d3a964a8faedd36d460&mode=3

regards

[Kapp Kapter]

Ah! good suggestions.

The server was brand new and running for 4 days and you are correct...I only received this error after the 4th day, which was my upgrade day.

That day, the Controller firmware was -not- upgraded but the Controller Driver, Hard Drive Firmware and the Storage Event Notification Service were upgraded.

This message has only appeared 3 times: 1st time about 30 minutes after the 1st reboot after all the upgrades were done. Then 3 days later it appear twice within the same second at a very quiet time for the server 5:16am

I did run the full offline HP Insight diagnostic on the drives last night and it came back clean.

I will scrape some time this weekend to try at least the Driver downgrade and the Cissesrv downgrade.

thx!

[KarloChacon]

yes sometimes installing latest updates give us new messages - most of the time - phantom messages so downgrading is the option

OK try that and let us know

bye

[Ed Roche]

what was the solution to this? I have just upgraded a dl380 g5 with proliant support pack 8.10 A after which I received this error. The P400 card fw remained the same (latest).
Is this error a problem?

[Kapp Kapter]

I was seeing this error about every other day after upgrading to lastest PSP and Firmware.

I only downgraded:

HP ProLiant Smart Array SAS/SATA Event Notification Service for Windows Server 2003/2008

(per Karlo's suggestion). I did this downgrade on Sept. 3 and have not seen the error since. I will post again in another couple of days as I watch the event log.

I did not have to reboot after this downgrade. The installation will stop and start the service.

[KarloChacon]

that is good


ok keep us posted

[MJT_1]

This event can occur following any unexpected power loss or server reset during host write operations. The inconsistency is only detected by Smart Array's background Surface Analysis (auto-reliability monitoring), and the event will continue to be reported on subsequent boots until the specified LBA range is ultimately overwritten by the host. Has the server ever been reset or lost power during write operations? Are the same LBA range(s) reported as being inconsistent each time the system is rebooted, or do they differ each boot?

[MJT_1]

Just a few more comments:

If the event always indicates a consistent LBA range and if there's no easy way to re-write this LBA range under the OS, another way to repair the inconsistency is to break and then re-sync the mirror pair (using offline ACU), or by manually hot-removing one of the drives from the mirror pair and reinserting (allowing it to rebuild).

Note that this event is more likely to be reported following an unexpected power loss/reset during writes if the controller's posted-write cache is disabled (such as in this case when there are no batteries). If the controller's battery-backed posted write cache was enabled, any write data held in the cache would automatically be written to BOTH drives after the power is restored, maintaining consistency between the mirrored drives.