Netserver RS/12 Netraid 3Si

[T_55]

I have a netserver RS/12 external rack array consisting of 8 drives with one of them a hot spare. This array is connected to a Netserver LHr4 server via a 3Si SCSI card. The onboard arrary is connected on channel 0 & 1 and the external array is connected via Channel 2 of the card.

Recently the hot spare failed and in addition at some point after that failure, disk #2 failed causing the drive to move it's info to the remaininng drives. Thus I am no longer running with RAID 5. When I replaced the 2 drives it appears that the autorebuild, which is enalbled, starts to work then it comes back with the drive failed. This happens with either of the new drives that I have purchased.

When I try a manual rebuild of the drive it gets to 5% and fails. So my question is do you think the problem lies with the SCSI card on Channel 2, or with the array itself. I replaced the scsi controller card that is on the array because we happened to have a second one on hand, but that gave no different results.

Any help would be greatly appreciated.

[Greg Carlson]

T,

If you are failing at 5% at the same time on multiple hdds, then one of your online hdds most likely has media errors. To verify this is the case, go into ctrl-M, objects and physical hdds and check for media errors and other errors and check all hdds. Most likely you'll find media or physical errors on one or more hdds. If this is the case, there are only two options available in this scenario for the netserver line.

First choice which is ugly is to replace the hdd which has media errors, which results in you losing more than one hdd in your RAID 5. (You are still running Raid 5, just in a degraded state)

Second choice will resolve your failed rebuild issue w/out failing a second hdd. Cable the RS/12 to a SCSI controller or move the hdd with media errors to a slot in the LH4 and enable embedded SCSI, then run a verify on the hdd(s) with any media errors and remap all bad blocks. Then you'll be able to rebuild the failed hdd.

Make sure you pull hdds from the RS/12 from a powered down state.

Ciao,
Greg

[T_55]

Thanks for the info, however, I have checked the media errors and others errors and everyone of them has no errors, which is the puzzling part.

Even after purchasing new drives, unless I have really bad luck and both drives are bad, that they both state that the rebuld fails even thought no errors.

This brings me back to could the SCSI card be the root of the problem.

[Greg Carlson]

T,

It is possible it is the controller. Do you have netraid assistant installed on your system? If you do, please post the netraid log for review. Are there any check conditions being reported? If you don't have netraid assist installed. Install it and then try another rebuild and then check the log for any check conditions.

Get netraid assist here:
http://h20000.www2.hp.com/bizsupport/TechSupport/DriverDownload.jsp?pnameOID=62475&locale=en_US&taskId=135&prodSeriesId=62473&prodTypeId=329290&swEnvOID=181

Greg

[T_55]

Is there a version Netware 6.0?

[Greg Carlson]

T,

This is the only utility available for Netware.
http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?locale=en_US〈=English&pnameOID=51917&prodSeriesId=51915&prodTypeId=329290&basePartNum=COL3579&locBasepartNum=ns-12646-1&os=Novell+NetWare+6.x&tech=Software++Storage+Controller&locale=en_US〈=English&pnameOID=51917&prodSeriesId=51915&prodTypeId=329290&basePartNum=COL3579&locBasepartNum=ns-12646-1&os=Novell+NetWare+6.x&tech=Software++Storage+Controller&attTypeId=110

I don't know if it creates a log file however with sense conditions. There must be somewhere in Netware where you can check for sense condition errors though.

Greg

[Marino Meloni_1 · Accepted solution]

I understand that you have a raid 5 with 8 drives plus hot spare.
you get two hdd failed (the hotspare and the drive 2)
now you are running a raid 5 in degraded mode (7 drives online plus drive 2 failed) (hotspare is still hotspare)
You checked the logs of all the drives but found no media errors, no other errors
rebuild of the raid with new drive in slot fail each time in slot 0
if this is your configuration, and the Greg suggestion do not work, probably you have a logical error on the array, this usualy is repaire when you run the "Consistency check" in background when the raid is redundant. Now, there is no way to repair from this.
I suggest you to prepare a full backup of your data.
clear the raid, and initialize it again, then restore the data.

The logical error on the raid may be present on one file (then one file will be corrupted) or in an area where there is no data

marino

[Greg Carlson]

T,

I just re-read your initial post and noticed you stated you had already tried another netraid card. So if none of your drives have any media errors, Marino is probably correct in assuming you have corruption in your RAID 5 stripe. A full restore of the array would be needed. Delete the array, reconfigure, initialize and restore from backup. Running scheduled consistency checks (not an option in degraded state) would potentially prevent this from happening again. Use the monitor tool in the earlier link to schedule weekly consistency checks. I think if memory serves correct, that they default to Sunday's (although you can change the time)

Greg

[T_55]

Thanks for you help. I have always figured that the end result was to blow away the array and rebuild it then restore from backup. Just didn't want to have to go that route unless necessary.