2400 RAID5 Problems

Shawn Neumann shawn at domain7.com
Wed Dec 4 01:48:00 CST 2002


I'm looking for some help on a raid issue on our PowerEdge 2400 that 
recently appeared.  I believe there are 2 separate issues, however I 
would like any advice available.  First is an apparent drive failure and 
secondly, our inability to access the afa0 controller.

We have a 5 disk RAID5 array and recently (as the log exerpts bellow 
show), it appears that our 3rd drive failed.

/messages.1:Nov 30 19:24:26 s2 kernel: aacraid:ID(0:03:0); Selection 
Timeout [command:0x28]
messages.1:Nov 30 19:24:27 s2 kernel: aacraid:Drive 0:3:0 returning error
messages.1:Nov 30 19:24:27 s2 kernel: aacraid:ID(0:03:0) - drive failure 
(retries exhausted)
messages.1:Nov 30 19:24:28 s2 kernel: aacraid:RAID5 Container 0 Drive 
0:3:0 Failure /

So now, the monday after the notification, we attempt to open afacli to 
examine the RAID. When running the "open afa0" command, we get one of 
the following 2 errors:

/The driver could not execute the requested IOCTL SENDFIB,  
25=Inappropriate ioctl for device. /

or simply:

/Command Error: <The requested controller does not exist.>

/The first error appears less frequently, however either one obviously 
prohibits us from continuing. It seems that it appearance of the first 
message is random rather than all attempts or ever 'x' number of 
attempts.  My question is if anyone is familiar with this series of 
events?  If so, what is the suggested course of action?  We can easily 
replace the failed 3rd drive, however I not sure if the RAID will be 
able to be rebuilt once we have done this since we cannot access afacli.

You thoughts are greatly appreciated!

We are running:
RedHat 7.3, afaapps-2.6-0, Linux version 2.4.18-18.7.xsmp
We recently updated the kernel (last week prior to the drive failure) 
and keep packages updated through RedHat's up2date network.



More information about the Linux-PowerEdge mailing list