LH3 Slow when booting

[Darrenrd]

Hi, All
I just recently install Win2000 SP3 along with the lates MS patches and also update my BIOS and flash the RAID controller on my LH3, howerver on bootup this server takes about 2 mins to check the memory on post and then about 3 mins staring the drives. I get the following message when the drives are being started "firmware starting drive" is this normal, I can't beleive that this server should take such a long time booting. Has anyone experience this problem before.. I'm running 2 RAID 5's with 6 9 GB on each array..
Thanks
Darren

[Martin Breidenbach]

I've seen this message before and there usually was a problem with one of the drives.

The drives don't start when power is supplied because they would draw to much power if they all would start simultaneously. So the RAID controller sends a 'start unit' command to each drive one at a time.

AFAIK the message "firmware starting drive" means that the drive takes a long time so spin up.

[Darrenrd]

Hi, Martin
I think you are correct by saying I might have a bad drive.. After reviewing my config I notice that I can only see 5 drives out of the six drives, however all the light are green on all six drives. Is there a way I can tell which drive is bad so I can replace it. On Dell server there is an option to blink drives, by right clicking the drives in the array, howerver with HP NetRaid I don't see an indication on how I can tell which drive is bad all it says is the SCSI ID number in which this case it is SCSI ID 15..

Thanks
-Darren

[Alicia White · Accepted solution]

Where are you viewing the drive configuration? The config doesn't show any of the drives as failed? What is the number of stripes for each array? Are you using 5 drives for each with one hotspare? Or, are you using all 6 drives as part of the RAID5 array on both?

On the tower model, the IDs on the LH3 are as follows for the right hand cage, from top to bottom):
ID 9
ID 8
ID 3
ID 2
ID 1
ID 0
Left drive cage, from top to bottom:
ID 15
ID 14
ID 13
ID 12
ID 11
ID 10

In a rack-mounted LH3, the SCSI IDs on the bottom drive cage are (from left to right):
ID 0, ID 1, ID 2, ID 3, ID 8, ID 9
On the top cage in a rack mount LH3, the SCSI IDs are(from left to right):
ID 10, ID 11, ID 12, ID 13, ID 14, ID 15

ID5 is reserved for the SAF-TE processor on each drive cage (it will probably show up as PROC in your config). ID7 is reserved for the SCSI or NetRAID controller.

Does this help you? In NetRAID Assistant you can spin up or down a drive to get the LED to flash, but I don't think you can do that in NetRAID Express Tools (the BIOS level utility).

And, I do agree that it does sound like a hard drive has failed. If your array doesn't show a failed hard drive (and therefore the array is in ONLINE and not DEGRADED status), then you probably have a hot spare drive that went bad.

Alicia

[Darrenrd]

Hi, Alicia
Tomorrow I will pull out the drive on SCSI ID 15. as per your email in this case the fail drive will be located in the left cage slot 1 from the top. Then I will try re-booting the server again. I recently aquired this server after someone recently install the OS, however it is configure with (2) RAID 5 . I can see all six drive in one bank with no Hot spare define, and only 5 drives on the second array. I think that it was orginally configured with six drive on this array as well with no hot spare defined. What seems a bit strange, is why the bad drive don't show up in the array . Should did drive be diplayed within the netraid manager ?

Thanks

[Marino Meloni_1]

Hi
in these case, you can identify the drive that do not spin up looking at the led on the drives at boot time:
when inizialisation of the netraid took place, the scsi controller send a command in order to start the drives one after the other.
on each drive you have two leds "power" and "activity"
the drive that have difficulty to spin-up ( it just may have difficulty an could be running after delay, do not mean he is in fault) will have the activity led lighted for more time that the other. That is a way to identify it.
Othe way is to install the agents, these will interact with biso managment and will light the red led on any failed disk

marino

[Terri Harris]

I wanted to be sure you are not seeing any errors during POST, since it does take at least 2 minutes before it even gets to the mass storage devices.

Is the System BIOS current? It should be 4.06.43PL, LH3 updates are here:

http://h20004.www2.hp.com/soar_rnotes/bsdmatrix/matrix50448en_US.html