Feature: Use NMX-C gRPC API for NVLink partition management#1580
Open
rtamma-nv wants to merge 23 commits into
Open
Feature: Use NMX-C gRPC API for NVLink partition management#1580rtamma-nv wants to merge 23 commits into
rtamma-nv wants to merge 23 commits into
Conversation
… instead of NMX-M. Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Add integration tests that communicate with a real nmx-c simulator rather than mock. These tests are not run unless RUN_NMXC_SIMULATOR_TESTS env variable is set. Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
…ulator Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
|
🌿 Preview your docs: https://nvidia-preview-pull-request-1580.docs.buildwithfern.com/infra-controller |
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
kensimon
requested changes
May 12, 2026
Contributor
kensimon
left a comment
kensimon
left a comment
There was a problem hiding this comment.
Haven't finished with the review yet, but I'd say the biggest problem right now is the breaking CLI changes, which don't immediately seem necessary to me.
kensimon
reviewed
May 12, 2026
kensimon
reviewed
May 13, 2026
kensimon
reviewed
May 13, 2026
kensimon
reviewed
May 13, 2026
rtamma-nv
force-pushed
the
feature/nmxc-paritioning
branch
from
24dda3f to
4d774b4
Compare
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
Signed-off-by: Roopesh Tamma <rtamma@nvidia.com>
rtamma-nv
force-pushed
the
feature/nmxc-paritioning
branch
from
4d774b4 to
065a4b7
Compare
Restore the separate commands. This will get added in another PR. Signed-off-by: Thomas McRoberts <tmcroberts@nvidia.com>
…oning Signed-off-by: Thomas McRoberts <tmcroberts@nvidia.com>
Signed-off-by: Thomas McRoberts <tmcroberts@nvidia.com>
tmcroberts97
approved these changes
May 15, 2026
tmcroberts97
approved these changes
May 15, 2026
kensimon
approved these changes
May 18, 2026
Contributor
kensimon
left a comment
kensimon
left a comment
There was a problem hiding this comment.
Nitpick is minor and can be addressed in a followup
Comment on lines
+68
to
+76
| impl Endpoint { | ||
| pub fn new(url: impl AsRef<str>) -> Result<Self, NmxcError> { | ||
| let uri = url | ||
| .as_ref() | ||
| .parse::<Uri>() | ||
| .map_err(|e| NmxcError::InvalidEndpoint(format!("{}: {e}", url.as_ref())))?; | ||
| Ok(Self { uri }) | ||
| } | ||
| } |
Contributor
There was a problem hiding this comment.
Nit: This can be impl FromStr for Endpoint if you want it to be a bit more idiomatic (let e: Endpoint = s.parse()?, etc)
Coco-Ben
approved these changes
May 18, 2026
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
5 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.
Description
Use NMX-C grpc API directly for managing NVLink partitions. Previously, core used to communicate with NMX-M for partition management. These changes remove all NMX-M support and replace it with NMX-C equivalent support.
There are no changes to tenant/user facing APIs for logical partition management or the general algorithm in how nvlink_partition_monitor applies tenant defined nvlink config and returns status/observations.
This PR brings in:
. new admin-cli command to add rack serial to NMX-C endpoint mappings
. libnmxc grpc client library for NMX-C
. new integration tests that support interacting with a real NMX-C running in simulator mode (in addition to mock)
. version 2.5 of nmx_c.proto
. replacement of all functionality provided by NMX-M rest API with NMX-C gRPC equivalent.
. other associated miscellaneous changes like nmxc_browser and metrics
Relevant NMX-C documentation changes will be part of a subsequent PR.
Type of Change
Related Issues (Optional)
Breaking Changes
Testing
Additional Notes