Server Down Raid5, 2disks status failure, 1 spare disk, please read.

What server model is this?  What RAID controller are we working with?  Did the drives go offline at the same time (possibly result of power outage or other event at that location), or did one drive die some time ago and the other just failed causing the array to go offline?  No guesses ... do you know for sure?

Hi, 

The server is a HP Proliant ML350 (no generation) .  The owner of the server is not one that likes active support
or maintenance done on the server. Due to holding back the costs.  The server is from march 2008, running even since.

Last time i was there to do a bit of servicing the server was about 6 months ago. At that time no problems with the harddisks.

Sorry Flash i wish i knew if they stopped working at once or seperately i dont know. They dont tell me if they would know, all seems
to work from the users pov.

They also turned down the offer for offlinw back-up in 2010. So there is no back-up. But since i carry a big heart i still would like to help them out.

If you dont want to i understand.  

no problems there,  thanks for replying anyway sofar Flash.

Kind regards

Ben

"Sorry Flash i wish i knew if they stopped working at once or seperately i dont know. They dont tell me if they would know, all seems
to work from the users pov."

That's ok ... it is important to know which one failed first, because the way you typically would attempt to repair this is to force online the LAST disk to fail (since the array becomes inactive at that point, its data will be an exact match for the array).  Forcing online a disk that had failed minutes, days, or even weeks ago would destroy the array data.

The only problem is that I don't have any experience with HP servers, so I can't say with any certainty that the methods used on Dell systems hold true for HP systems - although some RAID concepts will be universal, sometimes 'terminology' is as important to understand as the concept.

Analyzing controller or hardware logs, one might be able to discover when/which disk failed, THEN, IF that remaining disk is healthy (enough), it can be forced online to attempt to regain access to the array data.  THIS is where I won't be much help - obtaining information relative to HP hardware regarding the history and status of the disks in the array.  I simply am not familiar with the hardware (although there is a good chance that it uses an Adaptec or LSI controller similar to Dell controllers of the same era) or the software required to obtain the information/logs needed to discover what happened (for Dell servers, OMSA Live!, DRAC (depending on the system), or ESM (depending on the system) can be used - I'm not familiar with HP equivalents).

I would recommend calling HP or posting in their forums.  You might also consider posting at Experts-Exchange.com - I know there are some HP gurus over there.  Here is what I would do:  call HP and pay for the support incident to figure out your options.  Pass that cost on to your customer.  How do you sell it?  The truth:  Tell them their data is gone, and there is a very good chance it cannot be retrieved, but there is one more thing to try before sending the drives to a data recovery expert, which will cost a few thousand dollars (MUCH more than the cost of the Support Call).  Bottom line:  Your customer needs to know that, at this point, there are no free options - they lost the data, and the more they pay, the more likely it is they will get it back.

Feel free to follow up if there is anything else you'd like to know ... I'll provide what info I can.

Hi flash,   it worked, in a bit of strange manner but it did the job.  At this time too late to let you know how, but i will post it tomorrow.

Just to let you know.

Kind Regards,

Ben

Note: If you managed to revive the raid 5, only use this uptime window to back up your data (tape, USB drive or whatever) so you can destroy the raid 5 and recreate it with at least 3 known good drives. This should be your primary goal right now.