HPE Community read-only access December 15, 2018
This is a maintenance upgrade. You will be able to read articles and posts, but not post or reply.
Hours:
Dec 15, 4:00 am to 10:00 am UTC
Dec 14, 10:00 pm CST to Dec 15, 4:00 am CST
Dec 14, 8:00 pm PST to Dec 15, 2:00 am PST
ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

DL380 Failed to Start

 
neil barker
Occasional Advisor

DL380 Failed to Start

Anybody explain the following...

DL380 failed and would not restart. Red light on front indicating power problem. 2 power supplies removed and wored in another server - both fine. Power cables tried in another server - fine.

Server has 6 disks - system mirror and 4 data disks as a RAID10. Removed all 6 disks from server and press power button - server starts! Go thru process of illimination and find that one of the RAID10 disks is preventing the server from starting. Start server without this disk and then plug in once server is started and server crashes. Replace disk all is fine. Try 'faulty' disk in another server and that server fails.

Question: Why does a faulty disk prevent server from starting?
1 REPLY
Matti_Kurkela
Honored Contributor

Re: DL380 Failed to Start

Looks like the disk has a fault that will disturb traffic on SCSI bus. (The disk electronics might be short-circuiting a line or two of the bus to ground, or something like that.)

At least on DL380 G3, the SCSI cabling can be arranged in either "simplex" or "duplex" configuration.

If the simplex configuration is used, all the disks are on a single SCSI bus and a fault that blocks the bus will cause the entire bus to be inaccessible until the faulty disk is removed. In practice, this means the server will crash.

The duplex cabling configuration places 2 disks on one bus and 4 on another bus. In this configuration, a disk fault that causes bus problems will drop only 2 or 4 disks out of 6, depending on where the faulty disk is located.

If you want maximum reliability, try to place half of any mirror on one SCSI bus and the other half on another bus. If some fault incapacitates one whole bus, the mirror halves on the other bus will still allow the server to run.
MK