[SERVER-75362] Ticket holder monitor deadlock detection Created: 28/Mar/23  Updated: 05/Jan/24

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

Type: New Feature Priority: Major - P3
Reporter: Gregory Noma Assignee: Backlog - Storage Execution Team
Resolution: Unresolved Votes: 0
Labels: or-workload-management
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
Assigned Teams:
Storage Execution
Participants:

 Description   

Deadlocks in the system shouldn't happen, but occasionally some bugs have come up where we end up deadlocked due to ticket exhaustion. The ticket holder monitor should be able to detect such a scenario by observing that one of the ticket pools is exhausted along with zero throughput. If such a scenario is detected, we can attempt to remediate it by increasing the ticket pool size. Potentially this may allow the system to make progress and resolve the deadlock. Note that this would primarily be a fallback mechanism and thus we should still make sure that in our testing we are made aware that a deadlock scenario was reached.


Generated at Thu Feb 08 06:29:59 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.