[SOLR-14275] Policy calculations are very slow for large clusters and large operations

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Won't Fix
• Affects Version/s: 7.7.2, 8.4.1
• Component/s: AutoScaling
• Labels: scaling
• Assignee: Unassigned
• Reporter: Andrzej Bialecki
• Created: 21/Feb/20 00:26
• Updated: 25/Feb/21 09:58
• Resolved: 11/Nov/20 12:50

Description

Replica placement calculations performed during collection creation take extremely long time (several minutes) when using a large cluster and creating a large collection (eg. 1000 nodes, 500 shards, 4 replicas).

Profiling shows that most of the time is spent in Row.computeCacheIfAbsent, which probably doesn't reuse this cache as much as it should.

Issue Links

• links to: GitHub Pull Request #1329
• links to: GitHub Pull Request #1386
• mentioned in: Page Loading...
• mentioned in: Page Loading...