[rhn-users] Raid repair

[LONNIE TROTTER]

Has anyone experienced this. 

   We had a server setup with physical space for six scsi hard drives. 
The first three were constructed as a RAID1 array while the last three 
were constructed as a RAID5 array. One of the drives in the RAID5 array 
failed and was not repairable. The failed drive was removed and a 
replacement drive put in its place, we booted into linux rescue and 
fdsik'd the drive creating a partition table exactly the same as the 
drive that had failed. Went to boot back in to Enterprise to raidhotadd 
the replacement harddrive to the RAID5 array, but enterprise would not 
let us do that. The RAID5 array would not start because there were only 
two drives available, if the broken drive was placed back in the raid 
would come up in degraded mode. The problem here is that we had no 
physical space to have both the failed drive and the replacement drive in 
at the same time. This meant that the raid would not start, preventing us 
from raidhotadding the replacement drive. As a result we copied the data 
off of the raid while it was in degraded mode, put the replacement drive 
in and rebuilt the raid array, then copied the data back onto it. This 
was quite time consuming as the data we were moving around was around 25 
gigabytes. 

Lonnie Trotter
Network Administrator
Apache Junction Unified School District
Phone: 480-982-1110 Ext. 2050