02-25-2007
6,
0
Join Date: Jul 2006
Last Activity: 4 October 2009, 5:16 PM EDT
Location: UK
Posts: 6
Thanks Given: 0
Thanked 0 Times in 0 Posts
Hi,
Just saw this and thought I would offer my interpretation of ctds.
c#t#d#s# is the standard mark up of the disks or storage on a Solaris system.
As I understand it c = controller; t = target on that controller card; d = disk number; s = slice number on that disk.
From the metastat output you have shown it seems that:
You have 2 physical disks, c1t0d0 and c1t1d0, that have been partitioned into at
least 3 slices - s0, s1 and s3.
Each of these slices has been allocated to a metadevice (d10 and d20 in the case of s0) and then used to build a mirror - d0.
d0 is where you build the filesystem and mount the location - it just happens to be then sitting on a mirrored pair of slices.
Also, I forgot to say, there is no rule as to what metadevice number you use - it is not linked to the c#t#d#s# details. For example you could have d0 as d123 made up of d97 and d34. It really does not matter although it does help to perhaps have some kind of standard numbering convention. I think there is a max number of 128 metadevices as standard but this can be raised in the Solaris kernel perameters.
Hope that helps; I am still learning to so if there are anything incorrect then if someone else can be kind enough to let me know that would be cool.
Si