Problem with Cluster +RS12 +3Si Netraid

[Eli Abad]

Hello, I have a problem with a Cluster: 2xLH3000 with 3Si Netraid controller + 1xRS12 with 10 HD.
I have two arrays in the RS12, one RAID1 (2x9.1Gb) for the quorum and one RAID5 (8x36Gb) for Exchange data.
The soft is: W2000 SP2 + Microsoft Cluster + Exchange server 5.5 SP4.

The Exchange database has been corrupted, and when I execute the ESEUTIL /p /ispriv utility, a few minutes later, the Netraid of the server that has the disk control starts to beep.

Then the server doesn???t respond.

Can you help me?

The firmware of controllers and disks has been updated and software drivers for W2000 are the HP indicates.

Thanks

[Marco Hogeveen]

Eli,

Is the firmware of the 3si W.01.30?
Is clustermode enabled on both netraid controllers and do they have a unique SCSI ID? 6&7?
Try to run diagnostics on both netraids (in Express Tools, Ctrl-M during POST) to see if there??s a hardware error.
Is the netraid driver mraid35x.sys version 2.62?

What type of harddisks are you using?

Marco.

[Eli Abad]

Marco,

- The firmware of the 3si is W.01.30
- Clustermode is enabled on both netraid controllers
- Every Netraid controller has a unique SCSI ID: one the 6 and the other the 7.
- Results of diagnostics: All is correct.
- The netraid driver mraid35x.sys version is 2.62

- The model of HD are: D9416A, exactly HP04, with the firmware updated with the hp04ff33.lod file.

Thanks

[Marino Meloni_1]

hi
the beep indicate a hdd failure on the shared storage (because the integrated netraid didn't have beeper)
if the server hang it may be because you start netraid assistant and then after an inquire on the scsi bus, it hang on the defective hdd.
to check if you are in such a situation, look at the leds of the hdd in the rs/12
it any of the hdd have a red led, than this is the fault one and need to be replaced, if no one have red led, check if one of the hdd have the two green leds fixed lighted. this hdd is the faulty one.
removing this hdd, the netraid assistant will restart.
bye

marino

[Eli Abad]

When I restart the cluster, the shared storage has all hdd with green leds and I can see the activity.

When I execute eseutil /p /ispriv, there is a high activity on the disk but a few minutes later the 3Si controller starts to beep and the disk activity stops.

Then, every disk of the array has one green led.

I have run diagnostics from Express Tools and the result is that all is correct.

Now, I have restored the data from a backup to a stand-alone server. But I have saved a copy of corrupted database in order to check the shared storage.

Eli