Deadlock detection code can be stale

Bug #1887523 reported by Mohammed Naser on 2020-07-14
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
neutron
High
Unassigned

Bug Description

oslo.db has plenty of infrastructure for detecting deadlocks, however, it seems that at the moment, neutron has it's own implementation of it which is missing a bunch of deadlocks, causing issues when doing work at scale.

this bug is to track the work in refactoring all of this to use the native oslo retry.

Lajos Katona (lajos-katona) wrote :

I'm wondering if it should be an RFE, and discuss on drivers meeting, what do you think?

tags: added: db oslo
Changed in neutron:
importance: Undecided → High
status: New → In Progress

Hello:

What is described here is a very important problem and hits the core of the API, which is the DB. The patch proposed, only for quota, is a good example of the direction we should take.

As commented by Lajos, this should be commented in the drivers meeting along with https://review.opendev.org/#/c/739139/ (Loki testing for Neutron).

I think both, the testing FW class and the refactor, should be done closely in parallel.

Regards.

Slawek Kaplonski (slaweq) wrote :

I don't think that this should be an RFE. It's just about refactoring our code and use code from oslo instead of our (similar like with enginefacade IMO).
I think we can open Blueprint to track progress on it but I don't think we need to discuss anything related to it in the drivers team meeting.

Slawek Kaplonski (slaweq) wrote :

I just created BP: https://blueprints.launchpad.net/neutron/+spec/handle-deadlocks-in-oslo-way to track progress of this refactoring.

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers