I have two nodes "A" (primary) & "B". Each node have 3 resources. On node "B" i replace disks and after that i add each resource as secondary. Two resources successfully connect and sync, but with one resource i have issue. When on node "B" i stop this resource, node "A" shows
lv1 role:Primary
disk:UpToDate
b.host connection:Connecting
When i starts this resource on node "B", node "A" shows:
lv1 role:Primary
disk:UpToDate
b.host connection:StandAlone
and node "B" shows:
lv1 role:Secondary
disk:Inconsistent
a.host connection:Connecting
I try everything, remove/add resource, metadata on node "B", remove data, rebuild raid drbdadm connect lv1 --discard-my-data
etc.
One difference between worked an broken resources is This node was a crashed primary, and has not seen its peer since
+--< Current data generation UUID >-
| +--< Bitmap's base data generation UUID >-
| | +--< younger history UUID >-
| | | +-< older history >-
V V V V
E43824C7BC375B4A:626476078D91E933:CC1DC3FAD143EDCC:E4E71860FBA887C2:1:1:1:1:0:0:0:0:0:0:0:1
^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^
-< Data consistency flag >--+ | | | | | | | | | | |
-< Data was/is currently up-to-date >--+ | | | | | | | | | |
-< Node was/is currently primary >--+ | | | | | | | | |
-< This node was a crashed primary, and has not seen its peer since >--+ | | | | | | | |
-< The activity-log was applied, the disk can be attached >--+ | | | | | | |
-< The activity-log was disabled, peer is completely out of sync >--+ | | | | | |
-< This node was primary when it lost quorum >--+ | | | | |
-< Node was/is currently connected >--+ | | | |
-< The peer's disk was out-dated or inconsistent >--+ | | |
-< A fence policy other the dont-care was used >--+ | |
-< Node was in the progress of marking all blocks as out of sync >--+ |
-< At least once we saw this node with a backing device attached >--+
Any ideas how i can fix it?
UPD: Find new differences a kernel: drbd lv1/0 drbd1 b.host: The peer's disk size is too small! (999671944 < 1000196216 sectors)