[SERVER-74939] `CheckOrphansAreDeletedHelpers` may consider some owned documents as orphans Created: 16/Mar/23  Updated: 25/Apr/23  Resolved: 25/Apr/23

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: Marcos José Grillo Ramirez
Resolution: Duplicate Votes: 0
Labels: sharding-wfbf-day
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
duplicates SERVER-76343 modify stopBalancer to explicitly wai... Backlog
Assigned Teams:
Sharding EMEA
Operating System: ALL
Sprint: Sharding EMEA 2023-04-17, Sharding EMEA 2023-05-01
Participants:

 Description   

The hook checking that no orphaned documents are present on disk at the end of a test works as follows:

It has been observed a failure on step (3) because the hook wrongly believed that some documents actually owned by a shard were orphaned.

The reason of the failure was the following: shard1 was migrating data to shard0 and the hook started analyzing shard0:

  • Step 1 passed because the migration coordinator is only present on the donor side
  • Step 2 passed because the "pending" range deleter task document was removed right after the migration committed
  • Step 3 failed because the read from config.chunks did not account for the migration just committed, probably because reading from a secondary node of the config server


 Comments   
Comment by Marcos José Grillo Ramirez [ 25/Apr/23 ]

After having an internal discussion, the best solution is to make the stop balancer command to wait for outstanding migrations. That will be implemented in SERVER-76343, so, closing this ticket as a duplicate.

Generated at Thu Feb 08 06:28:55 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.