[CASSANDRA-17159] Log read timeouts and unavailables at INFO via a no-spam logger

• Type: Improvement
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 4.1-alpha1, 4.1
• Component/s: Observability/Logging
• Change Category: Operability
• Complexity: Normal
• Platform: All

Source Control Link

https://github.com/apache/cassandra/commit/33fd2dc817f56116af0b4b7d2cd5365ce52948d9

Test and Documentation Plan

• manual verification of logging output
• new in-jvm unit tests

• Assignee: Caleb Rackliffe
• Reporter: Caleb Rackliffe

Authors

Caleb Rackliffe

Reviewers

David Capwell, Yifan Cai

• Created: 17/Nov/21 18:53
• Updated: 27/May/22 19:24
• Resolved: 18/Nov/21 21:50

Description

We don’t have great logging visibility into timeouts and unavailables for reads and range reads. If we limit output to something like 1 per minute at INFO, we should be able to avoid spam but also have some record of what exactly failed. This could be especially helpful during node bounces and upgrades.