[HDDS-1610] applyTransaction failure should not be lost on restart

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Labels: pull-request-available
• Target Version/s: 0.5.0
• Assignee: Shashikant Banerjee
• Reporter: Shashikant Banerjee
• Created: 29/May/19 17:36
• Updated: 27/Aug/19 23:21
• Resolved: 20/Aug/19 09:18

Description

If the applyTransaction fails in the containerStateMachine, then the container should not accept new writes on restart,.

This can occur if

1. chunk write applyTransaction fails
2. container state update to UNHEALTHY also fails
3. Ratis snapshot is taken
4. Node restarts
5. container accepts new transactions

Issue Links

• is duplicated by: HDDS-1739 Handle Apply Transaction Failure in State Machine
• links to: GitHub Pull Request #1226