Etcd slaves smart start
Registered by
Alexander Tivelkov
Currently the etcd services do sleeps/delays on their start to prevent possibility of concurrent initialization.
This prevents split brain conditions, but is ugly and reduces deployment performance.
This should be optimized, so the services probe the status of etcd cluster before proceeding instead of simple unconditional waiting.
Blueprint information
- Status:
- Not started
- Approver:
- Serg Melikyan
- Priority:
- High
- Drafter:
- Alexander Tivelkov
- Direction:
- Approved
- Assignee:
- Alexey Deryugin
- Definition:
- Approved
- Series goal:
- Accepted for mitaka
- Implementation:
- Not started
- Milestone target:
- None
- Started by
- Completed by
Related branches
Related bugs
Sprints
Whiteboard
(?)
Work Items
Dependency tree
* Blueprints in grey have been implemented.