[SERVER-71259] Shard merge recipient might read donor file data after the backup cursor is closed. Created: 10/Nov/22  Updated: 05/Dec/22  Resolved: 05/Dec/22

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

Type: Bug Priority: Major - P3
Reporter: Suganthi Mani Assignee: [DO NOT USE] Backlog - Server Serverless (Inactive)
Resolution: Duplicate Votes: 0
Labels: shard-merge-milestone-3
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
depends on SERVER-71247 DocumentSourceBackupFile should ensur... Closed
Assigned Teams:
Serverless
Operating System: ALL
Participants:

 Description   

Currently DocumentSourceBackupFile cursor doesn't check if the backup cursor is alive in between batches, it only checks at the beginning of the command. So, merge recipients might read donor file data after the backup cursor is closed. For a restore using backup cursor works correctly only if the files were copied when the backup cursor is alive. In such scenarios, we expect running WT RTS  on such copied files might cause node crash/ WT instance to throw error. It's risky to always count on that WT behavior.

 



 Comments   
Comment by Suganthi Mani [ 05/Dec/22 ]

Since SERVER-71247 fixed the issue on the source side rather than target (i.e, DocumentSourceBackupFile will now ensure backup cursor is still open before returning data), we don't need to do any work for this ticket. So, closing this ticket as depends on SERVER-71247.

Generated at Thu Feb 08 06:18:29 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.