rebuild raid 5 using Adaptec Conf Utility

You can reboot the system and then go to the POST screen and type in the controller BIOS sequence (i think it is ALT-A, but can't remember) to go to the Adaptec card.

From there you can see whether or not it is rebuilding, and what disk is bad.  In general, I emphasize IN GENERAL, I would not expect it to rebuild anything because you had a drive failure.  You generally need to yank the bad drive and put another in its place and then it will rebuild.

Get a replacement drive, NOW, and in the interim, kick off a full backup. You have no room for error now.  If you lose a drive you have 100% data loss.  If you lose even 1 out of your half-billion blocks then you get partial data loss.

At this point the server is up and running 2 of the hard-disks are blinking green normally and the one which had yellow is now blinking green but at a different pace than the other ones.

1) What is the meaning of the blinking green on the apparent failured drive?
2) If I have to replace the hard-disk (which I can have monda) what is the process, should I shutdown the server, replace the drive and turning back on and go to the ACU menu, initilaize, etc.  OR I can just pull out the drive (without shuting down the server)  insert the new one and let it run?

The BIOS user guide might help - http://support.dell.com/support/edocs/storage/RAID/CS6CH/en/ug/dell_cee.htm#1122833
First step is look in the BIOS event log.

(it's on Dell's site since they use this card, HP don't use it but of course you can put whatever you want in the server within reason).

the problem is that documentation for a adaptec serial ata 2610SA using hp sata is not the best..very poor at least. Seems that this controller is not under the responsabillity of anybody, nor Adaptec neither HP.

Adaptec doesn't qualify disks with HP firmware.  But HP qualifies their OEM version of the adaptec controller with disks having their firmware.  So if you want something that is supported and tested, buy a different controller.

Sorry.

yes, the only way to access the tools it using ctral-alt during booting, but tha'st what I have but I'm sure I'm not the only case since this model was purchase as it is from HP.

Since you are running SBS2003, why not just make it a RAID1, and do it with the native windows software RAID1 (via dynamic disks).  Performance will be MUCH, MUCH better, and then you just disable the RAID hardware feature entirely.  

Then add another disk and make it a pair of RAID1s if you want the extra space.  Problem solved.

HP don't sell them wih Adaptec cards in them, OEM or not. A few of the older 100 series had Adaptecs but they weren't proper ProLiants - came from the HP systempro design team I think. Someone must have added it or it was bought remarketed.

Ok.
I got that server from HP with that card....hmm...or I got it from CDW?...it was 5 years ago and I'm not going to check the paperwork now

Regarding the important issue.  I got the hard disks working properly. I'm saying this only based on the fact that the 3 hdd have the green light flashing sync now. I did not reboot the system to check on the POST yet (production server) I'll do it on Saturday. I did not replace any hdd though.

Just to RECAP:

-update bios
-power cycle
-got amber light on one of the hdd
-went to OS and shutdown
-Remove and insert back the hdd with the amber light, turn the server back on.
-POST is saying rebuilding or damage, went to OS no problem
-The "damage" hdd is blinking green (regularly) but is not in-sync  wit the lights of the other 2 hdd
-The 3 hdd are now flashing green and in-sync. Did not reboot the system yet. Some users complain that is slow...but who knows....

My second question remain:
2) If I have to replace the hard-disk what is the process?
   -Should I shutdown the server, replace the drive and turning back on and go to the ACU menu, initialize, etc.
 - OR I can just pull out the drive (without shuting down the server)  insert the new one and let it run?

Thanks

Always replace disks HOT, while everything is powered up and running. That is what the controller expects you to do.   BUT best practice is to make sure the disk has been powered up before and burned in.  You don't want to put in a disk drive and watch it die 30 seconds into the process.

The Adaptec can't handle multiple failure scenarios like that .. so *poof* your data is gone.

Also best practice, Take full backup.   Make sure you do a full power off / power on before updating firmware as this clears out any memory leaks and does a full reset.

Thanks. I'll do a power cycle on Saturday to see what is the final state of this, I suppose I'll get 2 possible results

1) The failure amber light is back the raid1 is not optimal. If this is the case I'll replace (hot) the bad hardisk....
2) if the array is optimal I'll add another hdd and create spare....

Will see

I agree wholeheartedly with replacing hot if you can, at least then the controller has some idea that the new disk is really new and any metadata on it can be ignored.