[DRBD-user] Resize and snapshot a resource

Eduardo Grosclaude eduardo.grosclaude at gmail.com
Thu Feb 22 12:55:20 CET 2007

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


Hello, I'm looking for a way to have my DRBD resources resized and
snapshotted. I am testing the following DRBD over LVM setup and
procedures. Not sure if my thinking is sane as LVM still works as a
bit of a black box to my mind. Please criticize.

-I am using up to date DBRD 0.7 on CentOS 4.4 (2.6.9-42.0.8.ELsmp)
-Disk layouts are identical on both nodes
-Logical Volume V is into Volume Group G, some extra storage available in G
-DRBD resource R's "disk" attribute is V
-R's metadisk is on a separate partition, not a PV
-I do mkfs -t ext3 on R

-To resize R I plan to:
a) Add a new disk into secondary
b) Extend V at secondary with PV from new disk partition
c) drbdadm resize R at secondary
d) Repeat a, b, c at the other node
e) ext2online R at the primary, mounted

-To take a snapshot I plan to:
a') Create a snapshot S into G, from original V, at the primary
b') Use it, delete it

Is all this sound? So far, I have done a) above with a new partition
instead of a new disk, and e2fsck reports OK as I switch roles back
and forth. I have used no stripe/extent controlling features. Should I
care about other OK signs at the LVM level?
If alright, is there a better setup/procedure given these goals?
Thank you very much in advance
-- 
Eduardo Grosclaude
Universidad Nacional del Comahue
Neuquen, Argentina




More information about the drbd-user mailing list