[SERVER-70728] Migrations stall for `orphanCleanupDelaySecs` in mongod v4.4 with fcv mode v4.2 Created: 20/Oct/22  Updated: 30/Mar/23  Resolved: 17/Nov/22

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

Type: Bug Priority: Major - P3
Reporter: Pierlauro Sciarelli Assignee: Pierlauro Sciarelli
Resolution: Won't Fix Votes: 1
Labels: shardingemea-qw
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: File migrations_waiting_for_range_deleter.js    
Issue Links:
Related
Operating System: ALL
Sprint: Sharding EMEA 2022-11-14, Sharding EMEA 2022-11-28
Participants:
Case:
Story Points: 3

 Description   

[reproducible attached]

The recipient of a migration in v4.4 with FCV 4.2 always stalls for orphanCleanupDelaySec (excluding the very first migration, because the collection is not known).

The flow is the following: in case of FCV v4.2, the migration destination manager calls into beginReceive that unconditionally schedules a range deletion for the chunk being received waiting for the configured delay for secondary queries.

This results in all migrations stalling at least for orphanCleanupDelaySecs even when there is no orphaned document to remove.



 Comments   
Comment by Pierlauro Sciarelli [ 17/Nov/22 ]

Closing as won't fix because this can happen in a very small time window during upgrade/downgrade

Generated at Thu Feb 08 06:16:57 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.