Skip to content
This repository has been archived by the owner on Sep 30, 2024. It is now read-only.

graceful and forced failover: better anlaysis including GTID fix #509

Merged
merged 3 commits into from
May 21, 2018

Conversation

shlomi-noach
Copy link
Collaborator

Fixes #508

While this PR was originated to fix #508, it actually fixes more than that.

With graceful-master-takeover and force-master-failover, we impose a failover scenario on orchestrator even though it may not recognize one (indeed, in a graceful scenario there is no failure per se).

The analysis in those two scenarios was imposed on orchestrator. But as such, it lacked context and extra info. Namely, the one described in #508: lack of GTID context. Possibly also #421 is related.

With this PR we force the full analysis on the master on both these operations, and get the full context, including GTID state of the replicas.

cc @almeida-pythian @ManjotS

@shlomi-noach shlomi-noach changed the title graceful and forced failover: better nalaysis including GTID fix graceful and forced failover: better anlaysis including GTID fix May 20, 2018
@shlomi-noach shlomi-noach temporarily deployed to production/mysql_cluster=conductor May 21, 2018 08:40 Inactive
@shlomi-noach shlomi-noach merged commit ec2408f into master May 21, 2018
@shlomi-noach shlomi-noach deleted the graceful-takeover-auto-position branch May 21, 2018 10:37
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

MASTER_AUTO_POSITION being reset to 0 after graceful-master-takeover
1 participant