[SERVER-16540] Potential race in migration can have an aborted migration consume MigrateStatus::_cloneLocs - MongoDB Jira

XML

Word

Printable

JSON

Details

Type: Bug
Status: Closed
Priority: Major - P3
Resolution: Duplicate
Affects Version/s: 2.8.0-rc2
Fix Version/s: None
Component/s: Sharding
Labels:
None

Assigned Teams:

Sharding
Operating System:
ALL
Sprint:
Sharding 10 (02/19/16)

Description

There are cases where the donor shard can abort the migration without contacting the recipient shard. If it starts another migration before the recipient shard realizes it, then it can call _cloneMigrate against the donor shard and interfere with the new migration. A similar issue exist for _transferMods as well.

Attachments

Issue Links

duplicates

SERVER-20290 Recipient shard for migration can continue on retrieving data even after donor shard aborts

Closed

is related to

SERVER-16432 Chunk migration data structures not properly synchronized with storage engines that offer document-level locking

Closed

Activity

People

Assignee:: Backlog - Sharding Team

Reporter:: Randolph Tan

Participants:: Backlog - Sharding Team, Randolph Tan

Votes:: 0 Vote for this issue

Watchers:: 3 Start watching this issue

Dates

Created:: Dec 12 2014 09:59:03 PM UTC

Updated:: Dec 06 2022 04:58:07 AM UTC

Resolved:: Feb 01 2016 08:23:46 PM UTC