[DRBD-user] out-of-sync based on AL

[Cristian Zamfir]

Hi,

I am trying to understand why the AL is marking 196MB as out of sync.
I have two connected primaries A and B. After they are synced I detach 
B. In the mean time, the disk on A is not used. I reconnect B and the on 
disk bit map sees that no bytes are out-of-sync, but AL marks 196MB as 
out-of-sync.

In case I am misunderstandings something, can you please explain a bit 
how the AL works?

This is what dmesg says on node B:


[11989.600828] drbd1: disk( UpToDate -> Diskless )
[11994.007899] drbd1: disk( Diskless -> Attaching )
[11994.040025] drbd1: Found 6 transactions (50 active extents) in 
activity log.
[11994.040067] drbd1: max_segment_size ( = BIO size ) = 32768
[11994.050505] drbd1: reading of bitmap took 1 jiffies
[11994.050759] drbd1: recounting of set bits took additional 0 jiffies
[11994.050780] drbd1: 0 KB marked out-of-sync by on disk bit-map.
[11994.050819] drbd1: Marked additional 196 MB as out-of-sync based on AL.
[11994.066857] drbd1: disk( Attaching -> Negotiating )
[11994.066942] drbd1: Writing meta data super block now.
[11994.067332] drbd1: conn( Connected -> WFBitMapT )
[11994.067358] drbd1: Writing meta data super block now.
[11994.079356] drbd1: conn( WFBitMapT -> WFSyncUUID )
[11994.090853] drbd1: conn( WFSyncUUID -> SyncTarget ) disk( Negotiating 
-> Inconsistent )
[11994.090895] drbd1: Began resync as SyncTarget (will sync 200704 KB 
[50176 bits set]).
[11994.090926] drbd1: Writing meta data super block now.
[12014.611505] drbd1: Resync done (total 20 sec; paused 0 sec; 10032 K/sec)
[12014.611562] drbd1: conn( SyncTarget -> Connected ) disk( Inconsistent 
-> UpToDate )
[12014.611601] drbd1: Writing meta data super block now.


Thank you.

Cristian