[RFE] Use cached copy of hash ring in api service

Bug #1526451 reported by Vladyslav Drok
4
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Ironic
Fix Released
Wishlist
Zhenguo Niu

Bug Description

Currently, the API service resets the hash ring on every request, which would cause low performance in production environment. As we already cached the hash ring object, and recently introduced a way to refresh it during a interval, there could be some opportunities to reduce the db request.

Tags: needs-spec rfe
Vladyslav Drok (vdrok)
Changed in ironic:
status: New → Confirmed
importance: Undecided → Wishlist
tags: added: rfe
Vladyslav Drok (vdrok)
Changed in ironic:
assignee: nobody → Zhenguo Niu (niu-zglinux)
Revision history for this message
Jim Rollenhagen (jim-rollenhagen) wrote :
tags: added: needs-spec
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Change abandoned on ironic-specs (master)

Change abandoned by Zhenguo Niu (<email address hidden>) on branch: master
Review: https://review.opendev.org/250692

Revision history for this message
Julia Kreger (juliaashleykreger) wrote :

So, Ironic actually already does this. The challenge might be the process launch model, which could be a new process and a new copy every time. Regardless, I'm going to close this out as fixed release since it is already done, we can't account for every process launch model, and we sort of avoid the whole "who is authoritative maintainer" class of issues with a DLM or any other sort of recorder/tracker of the hash ring data/distribution.

Changed in ironic:
status: Confirmed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.