[CASSANDRA-10099] Improve concurrency in CompactionStrategyManager

• Type: Improvement
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 3.6
• Labels: compaction lcs
• Assignee: Marcus Eriksson
• Reporter: Yuki Morishita

Authors

Marcus Eriksson

Reviewers

Yuki Morishita

• Created: 17/Aug/15 18:06
• Updated: 16/Apr/19 09:31
• Resolved: 15/Mar/16 14:35

Description

Continue discussion from CASSANDRA-9882.

CompactionStrategyManager(WrappingCompactionStrategy for <3.0) tracks SSTable changes mainly for separating repaired / unrepaired SSTables (+ LCS manages level).

This is blocking operation, and can lead to block of flush etc. when determining next background task takes longer.

Explore the way to mitigate this concurrency issue.

Issue Links

• is duplicated by: CASSANDRA-10871 MemtableFlushWriter blocks and no flushing happens
• is duplicated by: CASSANDRA-10815 WrappingCompactionStrategy can block others when there are lots of sstables
• is related to: CASSANDRA-13948 Reload compaction strategies when JBOD disk boundary changes