ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

Drive failure, recovery...

Ayman Altounji
Valued Contributor

Drive failure, recovery...

greetings!

i am not a server person, nor a unix person...

we have a server (Proliant 3000) in the office running SCO Unix that crashed a while back due to a HDD that went bad. there are 3 x 9.1's in it. the support guy says that the system should not have crashed as the other 2 drives should have taken over and kept on truckin'. from another source i hear that the system needs 3 drives in order for the system to keep on running, so there needs to actually be 4 drives installed. this way, when one drive goes bad, the system will kick out the bad drive and include the spare (4th)into the system, thus enabling the system to recover and carry on working...

what is true in this case?

many thanx - john
1 REPLY
Ayman Altounji
Valued Contributor

Re: Drive failure, recovery...

Raid 5 was running on your system prior to the system being blitzed. the minimum number of drives you need for this form of fault tollerance is 3 hard disk drives, oh dear, nevermind, seriously think about adding an online spare for the system use and rebuild the server from backup. To accomplish building the online spare into your system use the Array Configuration Utility. Point to note, the online spare gives you presious time to replace the duff drive, don't look at it as being a saviour so think about building your system with 4 hard drives using raid 5 and also adding an online spare. bill in the post