Cross Pool Home-Away Preemption #3988
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* First step * Step 2 * Step 4 * Remove change * Remove change * Make nodedb and sctx upfront * Lint * WIP * wip - now compiling * Hack submitcheck to assign pools correctly * Minimal working version * Remove scheduling order * Fix tests * Lint * fix tests * Fix tests * Remove unused function * Add tests * Update submitcheck tests * Add ScheduledAway field to PodSchedulingContext * Add tests for scheduling_algo away scheduling * Remove unused testfixtures funcs * Remove debug code * Simplify pool config + add default config * Remove intersection * wip * Populate ResourceUsageByQueueAndPool on executor * Remove unused func * fix merge error * wip * Populate pool on pods + fix usage metric calculation * Simplify * Fix api tests * Update job_repository tests * fixed test * added extra test * lint * linting * remove stray file * Revert "remove stray file" This reverts commit faaaa969d6c4c93e933b316768c08650e178321a. * remove floatingResources for nodedb * fix test * renmae * rename to UnallocatableResources --------- Co-authored-by: chrismar503 <chris.martin@gresearch.co.uk>
* First step * Step 2 * Step 4 * Remove change * Remove change * Make nodedb and sctx upfront * Lint * WIP * wip - now compiling * Hack submitcheck to assign pools correctly * Minimal working version * Remove scheduling order * Fix tests * Lint * fix tests * Fix tests * Remove unused function * Add tests * Update submitcheck tests * Add ScheduledAway field to PodSchedulingContext * Add tests for scheduling_algo away scheduling * Remove unused testfixtures funcs * Remove debug code * Simplify pool config + add default config * Remove intersection * wip * Populate ResourceUsageByQueueAndPool on executor * Remove unused func * fix merge error * wip * Populate pool on pods + fix usage metric calculation * Simplify * Fix api tests * Update job_repository tests * fixed test * added extra test * lint * linting * remove stray file * Revert "remove stray file" This reverts commit faaaa969d6c4c93e933b316768c08650e178321a. * remove floatingResources for nodedb * fix test * renmae * rename to UnallocatableResources * fix test --------- Co-authored-by: jamesmur913 <james.murkin@gresearch.co.uk>
JamesMurkin
approved these changes
Oct 8, 2024
robertdavidsmith
approved these changes
Oct 8, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Implementation of cross-pool home-away preemption. Main Changes are as follows:
Pools now need to be enumerated up-front in the scheduler config, rather than dynamically determined.
Pools can now have one or more awayPools associated with them. For each awayPool, jobs may run as away jobs on that pool.
If a Pool only has awayPools set then it is essentially a synthetic pool. It contains no first-class nodes and jobs scheduled on it can only run as away jobs on other pools.
Synthetic pools work as normal pools, except their size is given as the capacity of the "away" pool(s) minus any home jobs running on those pools.