refactor(cuda): decompose CustomStorage helper - #119
Open
hhzhang16 wants to merge 1 commit into
Open
Conversation
hhzhang16
requested review from
Ronkahn21,
danbar2,
dfeigin-nv,
galletas1712,
hutm,
julienmancuso,
oleg-kushniriov,
shayasoolin and
shmuel-runai
as code owners
August 26, 2026 12:58
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. 🗂️ Base branches to auto review (1)
Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Enterprise Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
Signed-off-by: Hannah Zhang <hannahz@nvidia.com>
hhzhang16
force-pushed
the
hannahz/dep-1187-decompose-snapshot-cuda-checkpoint-helper-implementation
branch
from
August 27, 2026 14:15
9ac46f2 to
d86ee32
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Follow-up to #109 and #96 that decomposes the qualified CUDA CustomStorage helper; does not change its protocol or behavior. It does not need to merge in the initial CustomStorage code-freeze window.
Source ownership
flowchart LR CLI["main.cpp<br/>CLI dispatch"] --> Server["daemon_server.*<br/>socket, health, shutdown"] Server --> CUDA["cuda_operation.*<br/>driver operation + context lifetime"] CUDA --> Scheduler["transfer_scheduler.*<br/>workers, deadline, cancellation"] Scheduler --> Adapter["transfer_engine.*<br/>Snapshot-local NIXL POSIX adapter"]Main changes
main.cpp: reduced to CLI parsing and command dispatchdaemon_server: Unix-socket serving, health requests, shutdown, and RPC response handlingcuda_operation: CUDA initialization, operation completion, and target-scoped primary context lifetimetransfer_scheduler: per-extent worker creation, shared deadline, sibling cancellation, andresult aggregation
Preserves the existing cancellation test and adds scheduler empty-batch, invalid-result, and backend-failure coverage.
standalone and production Docker targets to build and test
No change to CUDA operation ordering, timeout behavior, unknown-outcome handling, target identity, or retained-context lifetime.
Exact-image legacy and POSIX Qwen3-0.6B checkpoint/restore/inference smoke pass