[DRBD-user] Cluster is giving me inode errors

[Sunil Varma]

Hello,

  I was using drbd version 0.7.10.

  My drbd resource is /dev/sda4 partition.
  After the first sync of the drbd I did fsck on the both the nodes 
(fsck /dev/sda4). It showed as clean.

  My cluster is working with combination of heart beat and drbd , When I 
did failover then in the primary node it was not able unmount the 
/dev/sda4 and it rebooted the primary node.
 After that secondary became primary .

  Then on the both nodes when I run fsck on /dev/sda4, it is showing 
i-nodes errors.\

  i-node errors are
--------------------------------------------------------------------------------------------------------------------- 

fsck 1.38 (30-Jun-2005)
e2fsck 1.38 (30-Jun-2005)
ext3 recovery flag is clear, but journal has data.
Run journal anyway<y>? yes

/dev/sdb4: recovering journal
Pass 1: Checking inodes, blocks, and sizes
Inode 8200225 is in use, but has dtime set.  Fix<y>? yes

Inode 8200225 has compression flag set on filesystem without compression 
support.  Clear<y>? yes

Inode 8200225, i_blocks is 1778545795, should be 0.  Fix<y>? yes

Inode 8200257 is in use, but has dtime set.  Fix<y>? yes

Inode 8200258 is in use, but has dtime set.  Fix<y>? yes

Inode 8200259 is in use, but has dtime set.  Fix<y>? yes
.
.
.
.
.
--------------------------------------------------------------------------------------------------------------------------

    
  Is i-node errors are due to hard reboot or any thing related old drbd 
version?
  Will these i-node errors will get copy to the other node also during 
the sync process ?

  Thanks in advance.