[SERVER-22233] Reduce the secondary throttling during chunk migration Created: 19/Jan/16  Updated: 19/Apr/18  Resolved: 21/Apr/16

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

Type: Task Priority: Major - P3
Reporter: Kaloian Manassiev Assignee: Judah Schvimer
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
is depended on by SERVER-22562 Protocol Version 1 causes very very s... Closed
Duplicate
is duplicated by SERVER-23340 Turn off moveChunk secondaryThrottle ... Closed
Gantt Dependency
Related
related to SERVER-22769 cleanupOrphaned has duplicated writeC... Closed
is related to SERVER-22276 implement "j" flag in write concern a... Closed
Backwards Compatibility: Fully Compatible
Sprint: Repl 14 (05/13/16)
Participants:

 Description   

This ticket combines the work to evaluate the performance impact of reducing (or even dropping) the secondary throttle during chunk migration and the actual code changes to make no throttle default.

This affects both the inserts and deletes – it might be different issues wrt/write concern.

The easiest change we can make is to substitute the w:2 writes after each document with a single majority write at the end of each stage.

These changes may or may not be feasible for MMAP V1, but given that as of version 3.0 the collection locks are fair, it might be possible to reduce the throttling for MMAP V1 as well.


Generated at Thu Feb 08 03:59:48 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.