feat(gpu_memory): support milligpu allocation logic#1131
Open
JustinChengLZ wants to merge 16 commits intokubewharf:mainfrom
Open
feat(gpu_memory): support milligpu allocation logic#1131JustinChengLZ wants to merge 16 commits intokubewharf:mainfrom
JustinChengLZ wants to merge 16 commits intokubewharf:mainfrom
Conversation
JustinChengLZ
requested review from
cheney-lin,
luomingmeng,
sun-yuliang,
waynepeking348 and
xu282934741
as code owners
JustinChengLZ
force-pushed
the
dev/milligpu-allocation
branch
2 times, most recently
from
e639157 to
040bb88
Compare
JustinChengLZ
force-pushed
the
dev/milligpu-allocation
branch
5 times, most recently
from
f859476 to
a31b04a
Compare
luomingmeng
force-pushed
the
dev/milligpu-allocation
branch
2 times, most recently
from
f4fb8a0 to
572ddda
Compare
Add DiscoverMemoryTopology function to calculate normal memory capacity by excluding static hugepages. This enables proper memory accounting in NUMA systems with hugepages configured. The function is now used by the existing Discover method to maintain consistent memory topology information.
Update memory calculations to use NormalMemoryCapacity and NormalMemoryDetails instead of total memory capacity to exclude static hugepages. This provides more accurate resource upper bounds and per-NUMA capacity calculations for memory provisioning and headroom estimation.
Add StaticHugePagesDetails and StaticHugePagesCapacity fields to MemoryTopology to track static hugepages allocation per NUMA node and system-wide. Update test cases and dummy topology generation to include these new fields.
Static huge pages should not be included in the total available memory calculation as they are reserved and cannot be used for dynamic allocation. This change updates the policy to properly account for static huge pages and adds corresponding test cases to verify the behavior.
…lation Add GetMemoryTopology interface to ReadonlyState and implement memory topology support in state management. This enables more precise memory capacity calculations by considering NormalMemoryDetails while excluding hugepages. The change affects multiple state generation functions across the codebase to properly handle memory topology information.
feat(qrm-plugin): support reclaimed cores vpa
luomingmeng
force-pushed
the
dev/milligpu-allocation
branch
2 times, most recently
from
493bb3e to
949d240
Compare
…cation logic - Rename all fractional GPU references to virtual GPU for consistency - Optimize topology hint calculation with pre-computed resource contexts - Improve NUMA node selection logic for packing/spreading strategies - Update environment variable names and config fields to reflect virtual GPU terminology - Add comprehensive tests for spreading mode and edge cases
…andling and logging Add input validation and detailed logging for device allocation process Include pod information in logs for better traceability
luomingmeng
force-pushed
the
dev/milligpu-allocation
branch
from
949d240 to
c5a48f7
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
What type of PR is this?
What this PR does / why we need it:
Which issue(s) this PR fixes:
Special notes for your reviewer: