Feature #18113
closed
[a-d-c] non-zero defaults for MaxCloudOpsPerSecond and MaxConcurrentInstanceCreateOps
Added by Tom Clegg over 2 years ago.
Updated over 2 years ago.
Release relationship:
Auto
Description
Instead of 0 (unlimited) we could start with values that are reasonable for a normal size production cluster / typical cloud account.
MaxConcurrentInstanceCreateOps was invented to accommodate Azure limitations, but now that we have it, we might as well promote using it as a guard rail for all cloud providers.
Perhaps 10 ops per second and 1 concurrent create op. Recommend raising to 20 concurrent create ops for Azure since it doesn't return the "create" response until the instance has booted.
- Tracker changed from Idea to Feature
- Description updated (diff)
- Related to Bug #18102: max dispatch attempts error added
- Description updated (diff)
- Target version set to 2021-10-13 sprint
- Assigned To set to Ward Vandewege
- Status changed from New to In Progress
The two comments on MaxConcurrentInstanceCreateOps read a bit strange. Perhaps the "recommended by Azure" & azure docs link should move to the 2nd paragraph and the "can be used with other cloud providers too, if desired" bit in the first paragraph can be removed entirely?
The old comment says "node creation", while we're here I think it would be more consistent to change that to "instance creation".
Besides that, LGTM
Ward Vandewege wrote:
Ready for review in 1e8b4a7843b63a418aca83e8dd0a24c0ce52a522 on branch 18113-change-cloudops-defaults.
Tests running at developer-run-tests: #2717 
I know Tom took the review but I saw the commit message go past and I looked at it.
The comments don't really explain the difference/relationship between MaxCloudOpsPerSecond and MaxConcurrentInstanceCreateOps.
Yes, I think that addresses all concerns. LGTM.
- % Done changed from 0 to 100
- Status changed from In Progress to Resolved
Also available in: Atom
PDF
Updated by 
Updated by 
Updated by