Sorry for the late reply. Just now finding this discussion. Here are my notes on how to do this. Skips step 4 & 5 if not using LVM. Skip step 3 if not using devmapper mulitpathing.
1) Must grow the disks on the SAN array first
2) Must have Linux recognize the disks with a device rescan
echo 1 > /sys/block/sdX/device/rescan
For multipath devices, find all sd?? devices corresponding to this mpath device (use ‘multipath -ll’ command), and do this rescan for all of them:
for i in sdao sdas sdaw sde
do
echo 1 > /sys/block/${i}/device/rescan
done
3) Must have device-mapper update it’s internal tables for this multipath device
multipathd -k"resize map <multipath_device>"
<multipath_device> can be “mpath14” or a dm WWID
Older versions of device-mapper-multipath require a manual table reload. See the dmsetup command (dmsetup table, dmsetup suspend, dmsetup reload, dmsetup resume).
**4) Resize the PVs
**
pvresize /dev/dm-XX
Run ‘pvs’ command before and after to see if the PV has grown.
5) Extend the logical volume
lvextend -l 100%VG -v /dev/vg01/lvol01
**6) Resize the file system
**
resize2fs /dev/vg01/lvol01
or
resize2fs /dev/mapper/mpathXX
a. command form is resize2fs <device> <size>
b. <size> can have K, M or G, or can be left out to allocate all space
c. can grow ext3 fs online
d. can shrink, but fs must be dismounted
e. command will take a long time for multi-TB file systems
f. you can watch it grow with the df -h command (once the resize command says “Performing an on-line resize…”)