If a participant with the highest zxid (lets call it A) isn't present during leader election, a participant with a lower zxid (say B) might be chosen as a leader. When A comes up, it will replay the log with that higher zxid. The change that was in that higher zxid will only be visible to the clients connecting to the participant A, but not to other participants.
I was able to reproduce this problem by
1. connect debugger to B and C and suspend them, so they don't write anything
2. Issue an update to the leader A.
3. After a few seconds, crash all servers (A,B,C)
4. Start B and C, let the leader election take place
5. Start A.
6. You will find that the update done in step 2 is visible on A but not on B,C, hence the inconsistency.
Below is a more detailed analysis of what is happening in the code.
1. Lets say there are three nodes in the ensemble A,B,C with A being the leader
2. The current epoch is 7.
3. For simplicity of the example, lets say zxid is a two digit number, with epoch being the first digit.
4. The zxid is 73
5. All the nodes have seen the change 73 and have persistently logged it.
Request with zxid 74 is issued. The leader A writes it to the log but there is a crash of the entire ensemble and B,C never write the change 74 to their log.
B,C restart, A is still down
B,C form the quorum
B is the new leader. Lets say B minCommitLog is 71 and maxCommitLog is 73
epoch is now 8, zxid is 80
Request with zxid 81 is successful. On B, minCommitLog is now 71, maxCommitLog is 81
A starts up. It applies the change in request with zxid 74 to its in-memory data tree
A contacts B to registerAsFollower and provides 74 as its ZxId
Since 71<=74<=81, B decides to send A the diff. B will send to A the proposal 81.
The problem with the above sequence is that A's data tree has the update from request 74, which is not correct. Before getting the proposals 81, A should have received a trunc to 73. I don't see that in the code. If the maxCommitLog on B hadn't bumped to 81 but had stayed at 73, that case seems to be fine.
|Field||Original Value||New Value|
|Priority||Critical [ 2 ]||Blocker [ 1 ]|
|Status||Patch Available [ 10002 ]||Open [ 1 ]|
|Assignee||Vishal Kathuria [ vishal.k ]|
|Fix Version/s||3.3.4 [ 12316276 ]|
|Status||Patch Available [ 10002 ]||Resolved [ 5 ]|
|Resolution||Fixed [ 1 ]|
|Status||Resolved [ 5 ]||Closed [ 6 ]|