[DRBD-user] block drbd1: The peer's disk size is too small!

Thomas Roth t.roth at gsi.de
Wed Aug 26 19:36:36 CEST 2009

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

Hi all,

I have this experimental setup where machine A acted as master, machine
B as slave, working fine. Then I disconnected and put the slave B into a
3-node setup, now acting as master. This also worked.

Now I went back to the original setup, trying to connect A (as master)
with B (as slave). Didn't work, and I got
> block drbd1: The peer's disk size is too small!

This is no real surprise, in the course of doing the 3-node setup I put
drbd-metatdata onto the partitions, so it would appear smaller
afterwards - at least that's what I thought.

On the level of "df", both partitions are identical of course.

However, I had a look at /proc/partitions and found that
858337640 drbd1     old and new master
858372184 drbd1     3-node-metatdata-laden slave:  larger !

Can I assume that putting the additional metadata on that slave took
more than the 34544 blocks it is larger than its peer?
And is there a way to calculate the space needed?


More information about the drbd-user mailing list