This repository has been archived by the owner on Sep 30, 2024. It is now read-only.
Orchestrator GUI incorrectly shows recovery option for intermediate database in chained replication #1463
Comments
kamil-holubicki
added a commit
to kamil-holubicki/orchestrator-openark-fork
that referenced
this issue
Dec 29, 2022
intermediate database in chained replication openark#1463 Problem: If we've got replication chain A->B->C, and C is down, GUI shows 'Recover' dropdown for node B, but there is no possible recovery action available in such a case. Cause: The root cause of the problem is the analysis logic in Analysis_dao.go:GetReplicationAnalysis(). The condition for setting AllIntermediateMasterReplicasNotReplicating does not check if there are any replicas reachable. So the case when all replicas are dead (no recovery action possible) and the case when some replicas are still reachable, but are not replicating (recovery action possible) are undistingushable. Solution: Improve the analysis logic. Report AllIntermediateMasterReplicasNotReplicating only if all replicas are not replicating, but there are still some reachable replicas. This commit also contains improvement of not trying to query the node which is not reachable (ping node before examinig it)
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
In a chained replication environment such as A -> B -> C. The loss of the leaf node, C, makes Orchestrator GUI show the Recover button on B when no action is possible. This can be confusing for users unfamiliar with MySQL topology and may think there is a possible action.
To reproduce this issue, you can create a cluster using 3 nodes with anydbver:
./anydbver deploy hn:ps0 ps:5.7 node1 hn:ps1 ps:5.7 master:default node2 hn:ps2 ps:5.7 master:node1 node3 hn:orc orchestrator master:default
Shutdown the ps2 instance and access the GUI. It will show the instance ps1 to recover.
The text was updated successfully, but these errors were encountered: