[CASSANDRA-15954] When compaction gets interrupted, the exception should include the compactionId

• Type: Improvement
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 4.0-beta3, 4.0
• Component/s: Local/Compaction
• Change Category: Operability
• Complexity: Low Hanging Fruit
• Platform: All

Source Control Link

https://github.com/apache/cassandra/commit/462261f4a495ec9be62091b26a7aa72470d95fda

Test and Documentation Plan

unit test

• Assignee: Yifan Cai
• Reporter: David Capwell

Authors

Yifan Cai

Reviewers

Caleb Rackliffe, David Capwell

• Created: 17/Jul/20 00:18
• Updated: 03/Jan/21 17:02
• Resolved: 08/Sep/20 23:49

Description

When we log in compaction we use the compactionId (or taskId), because of this we can figure out the start and end log lines for compaction. The issue is, when you interrupt a compaction task, we don’t log the compactionId, we instead log the tableId; for this reason it is much harder to attribute the interrupt to a single compactionId

Examples

INFO  2020-07-15T18:04:51,670 [CompactionExecutor:144057] org.apache.cassandra.db.compaction.CompactionTask:158 - Compacting (5a463760-c700-11ea-8982-13c71a558319) [data/ks/table/ma-27-Data.db:level=0, data/ks/table/
ma-24-Data.db:level=0, ]
INFO  2020-07-15T18:33:47,550 [CompactionExecutor:144057] org.apache.cassandra.concurrent.DebuggableThreadPoolExecutor:1942 - Compaction interrupted: Compaction@057aa994-c35b-39ec-b74d-33fba2d13bbc(ks, table, 3105436904/8658590553)bytes