[BBLISA] RAID Array Not Recognizing All 7TB

[Richard 'Doc' Kinne]

Hi Folks:

I'm a new, local SysAdmin who just moved into the area. I have  
experience SysAdmin-ing for some years, but because of the market,  
most of it was Windows based. I have some experience under Linux, but  
I need more to make me comfortable. I'm working at that.  :-)

I've been monitoring the list for some time, and now hope to become a  
bit more active now that I'm finally actually working in Cambridge.

Along those lines, I'm hoping that someone might have some clue on a  
recent issue I've had.

We recently upgraded a SATA RAID-5 array on one of our servers. We  
installed 8 1TB SATA drives in the system expecting to get a 7TB RAID  
array.

The server is running Fedora Core 6. The kernel is at 2.6.22.9 so we  
don't have a problem with respect to space limits. Initially when we  
initialized the RAID-5 array we set up 7 disks in the array with one  
as a hot-swap. We realized later that was not what we wanted to do,  
and thought that this might have been the reason why we were only  
getting 6TB when we thought we should be getting 7TB.

We went back and believe we "expanded an existing array" so that it  
brought in the 8th disk. Once that was done it started re- 
initializing the array in the background, which we thought was fine.  
We left it to its devices and took up the job the next morning.

The next morning, after the array was apparently re-initilized we  
used "parted" to manipulate the array. It still said that it only  
recognized 6TB, not 7TB. :-( We:

mklabel gpt
mkpart 0% 100%
print

and it still said it only recognized 6TB, not 7TB.

This seemed to be actual because I followed through by formatting the  
partition to the ext3 file system and mounting it. it showed 6TB  
ready to go.

As far as I know the RAID controller can handle 32TB, as can the  
kernel. The only thing I can think of at this point is that we need  
to start the whole thing from scratch, i.e. don't "expand an existing  
array," but re-do the array initialization completely.

Does anyone have any thoughts?  Many thanks.

---
Richard 'Doc' Kinne, [KQR]
American Association of Variable Star Observers
<rkinne @ aavso.org>



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.bblisa.org/pipermail/bblisa/attachments/20080827/8b3c53f9/attachment.htm