[SERVER-63244] Majority reads must round their read timestamp up to the oldest timestamp Created: 03/Feb/22  Updated: 29/Oct/23  Resolved: 11/Feb/22

Status: Closed
Project: Core Server
Component/s: None
Affects Version/s: None
Fix Version/s: 5.3.0

Type: Bug Priority: Major - P3
Reporter: Louis Williams Assignee: Daniel Gomez Ferro
Resolution: Fixed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
Problem/Incident
is caused by SERVER-55030 Remove mutexes that serialize seconda... Closed
Backwards Compatibility: Fully Compatible
Operating System: ALL
Sprint: Execution Team 2022-02-21
Participants:
Linked BF Score: 160

 Description   

SERVER-55030 reduced the critical section of a mutex that prevented the oldest_timestamp from advancing past the majority-commit point when starting new reads with the kMajorityCommited ReadSource (i.e. readConcern: majority).

There is a race now where the oldest timestamp can advance past the majority commit point before the read timestamp is set, resulting in a fassert.

We should use the "roundup_timestamps=(read=true)" configuration when starting these transactions and then return the actual read timestamp by querying WT.



 Comments   
Comment by Githook User [ 11/Feb/22 ]

Author:

{'name': 'Daniel Gómez Ferro', 'email': 'daniel.gomezferro@mongodb.com', 'username': 'dgomezferro'}

Message: SERVER-63244 Round up read timestamp for majority reads
Branch: master
https://github.com/mongodb/mongo/commit/89600c35511ce81a62ae235b27746a0698bdfef2

Generated at Thu Feb 08 05:57:15 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.