[SERVER-27798] Improve confusing Balancer log messages Created: 24/Jan/17  Updated: 06/Dec/22  Resolved: 13/Nov/17

Status: Closed
Project: Core Server
Component/s: Sharding
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Dianna Hohensee (Inactive) Assignee: [DO NOT USE] Backlog - Sharding Team
Resolution: Won't Fix Votes: 1
Labels: neweng
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Assigned Teams:
Sharding
Participants:

 Description   

When you have a new primary, it starts up the balancer thread. Before starting regular balancing, it tries to recover any migrations that were abandoned when the previous primary stepped down. But the logging for this is poor.

The message "CSRS balancer thread is recovered" (in balancer.cpp) is confusing. Particularly when it doesn't actually recover anything, or it's the first term of the replica set and there was no previous primary. What is being recovered is never explained.

Perhaps adding counters to the recovery functions so that we know whether anything happened, and explaining what 'recovered' or 'recovering' actually mean.


Generated at Thu Feb 08 04:16:16 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.