[HDDS-1859] Need a way to determine relative position of OM instances when cluster is down

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Information Provided
• Component/s: OM HA, Ozone Manager
• Target Version/s: 0.5.0
• Assignee: Unassigned
• Reporter: Arpit Agarwal
• Created: 24/Jul/19 22:36
• Updated: 24/Feb/22 20:16
• Resolved: 24/Feb/22 20:16

Description

It will be useful to figure out the relative positions of the OM instances when the cluster is down.

In HDFS world, we can do this by examining the edit log files and the seen_txid file.

In Ozone, this is slightly more difficult. We have to determine

1. relative positions of Ratis logs, i.e. log index and committed index. We may need to build some offline Ratis log parser for this.
2. relative positions of state machines. i.e. last applied index.