Project

General

Profile

Actions

Feature #20601

closed

Optimize arvados-dispatch-cloud performance when queue is very large

Added by Tom Clegg about 1 year ago. Updated 11 months ago.

Status:
Resolved
Priority:
Normal
Assigned To:
Category:
-
Story points:
-
Release relationship:
Auto

Description

Possible improvements include
  • When fetching queued containers, sort by priority, and stop when reaching expected capacity
  • Un-select mounts field when re-fetching a container whose mounts field we already have
  • Re-fetch container records less frequently when busy (but still frequently check for new queued containers if we have capacity to run them)

Subtasks 1 (0 open1 closed)

Task #20628: Review 20601-big-ctr-queueResolvedTom Clegg06/13/2023Actions
Actions

Also available in: Atom PDF