Subj : OS/2 Raid Controller...
To   : Peter Knapper
From : Oliver Rick
Date : Wed Mar 02 2005 06:56 pm

Hello, Peter!

On Mon, 28 Feb 2005 Peter Knapper wrote to Oliver Rick:

PK> BTW, judging by the datestamps on your message it took about 6 days for
PK> your reply to get here... must be a waystation along the path
PK> somewhere......;-)

I only poll every now and then, maybe once or twice a week. So when I write a 
reply, it lays around for some days before it will go out to my boss node.

If you want to speed things up a bit, you might like to move the subject to: 
news:comp.os.os2.setup.storage
I know that there are people who have actual experience with RAID HBAs under 
OS/2.

OR>> Well, then it's no longer "minimal interruption", I suppose? ;-)

PK> I consider a 15min outage to power down, change an HD, then power up again,
PK> with NO data loss, to be pretty insignificant compared with several hours
PK> to replace a failed drive, partition and restore data, but then I guess I
PK> am easy to please in this regard.......;-) The catch with replacement
PK> drives is that they are always bigger than before and revised partitioning
PK> decisions need to be made.......;-) I guess I like a simpler style of life,
PK> as long as it does things MY way......;-)

OR>> That still makes 3 drives for a RAID 5 or dumb RAID 1
OR>> mirroring with 2 drives.

PK> The 3 drive option seems like the most suitable one for me...

I don't know about the features of the NetCell, but with RAID 5 you'll have to 
go with SCSI. But after the replacement of a failed disk the hostadapter will 
need to rebuild the RAID which will take more than the 15 minutes you mentioned 
above with larger sizes, I'm afraid, since the hostadapter needs to *calculate* 
the lost data from the remaining data. However, IIRC, you can still use the 
array during rebuild, but it will be slow.

   /Olli/
--- GoldED/2 3.0.1
 * Origin: http://www.warpupdates.mynetcologne.de/ (2:2450/502.10)

.