Fix PrepareResourceClaim to read the rdmadev from sysfs path as well by neaggarwMS · Pull Request #57 · kubernetes-sigs/dranet

neaggarwMS · 2026-01-28T00:23:20Z

his PR fixes the PrepareResourceClaim function to read the RDMA device name from sysfs path as a fallback when rdmamap fails to detect the device.

Changes:

Added `getRdmaDeviceFromNetdev` function in `rdmadevice.go` that:
- First tries to get RDMA device using `rdmamap.GetRdmaDeviceForNetdevice()`
- Falls back to reading from sysfs path `/sys/class/net//device/infiniband/` if rdmamap fails
- This addresses a known issue with detecting RDMA devices for certain Mellanox NICs: IsRDmaDeviceForNetdevice() incorrectly returns false for InfiniBand (IPoIB) interfaces due to node GUID vs port GUID mismatch Mellanox/rdmamap#15
Added comprehensive unit tests in `rdmadevice_test.go`

Testing:

Unit tests pass
Tested on AKS cluster with Azure RDMA NIC (SKU: Standard_ND96isr_H100_v5)

k8s-ci-robot · 2026-01-28T00:23:27Z

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: neaggarwMS
Once this PR has been reviewed and has the lgtm label, please assign michaelasp for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

linux-foundation-easycla · 2026-01-28T00:23:28Z

❌ - login: @neaggarwMS / name: Neha Aggarwal . The commit (f15af9c) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please submit a support request ticket.

k8s-ci-robot · 2026-01-28T00:23:29Z

Welcome @neaggarwMS!

It looks like this is your first PR to kubernetes-sigs/dranet 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes-sigs/dranet has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

k8s-ci-robot · 2026-01-28T00:23:30Z

Hi @neaggarwMS. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

tamilmani1989 · 2026-01-28T20:48:02Z

pkg/driver/rdmadevice.go

+
+		//Fallback to sysfs check if rdmamap fails. This is particularly related to a known
+		// issue to detect RDMA devices for certain Mellanox NICs
+		// https://github.com/Mellanox/rdmamap/issues/15
+
+		rdmaDir := filepath.Join("/sys/class/net", ifName, "device/infiniband")
+
+		entries, err := os.ReadDir(rdmaDir)
+		if err != nil {
+			return "", fmt.Errorf("no RDMA device for %s: %w", ifName, err)
+		}
+
+		for _, entry := range entries {
+			if entry.IsDir() {
+				return entry.Name(), nil // Return first RDMA device found (e.g., "mlx5_0")
+			}
+		}
+
+		return "", fmt.Errorf("no RDMA device found for %s", ifName)
+	}
+
+	return rdmaDev, nil


should we consider creating this as function in pkg/inventory/sysfs.go as all sysnet operations added in that file

+1

Extending this further, can we please refactor the existing function

dranet/pkg/inventory/sysfs.go

Line 112 in 97d40e9

func hasRDMADeviceInSysfs(ifName string) bool {

into two functions, similar to the upstream implementation of IsRDmaDeviceForNetdevice and GetRdmaDeviceForNetdevice, where IsRDmaDeviceForNetdevice is a simple wrapper over GetRdmaDeviceForNetdevice (Ref. https://github.com/Mellanox/rdmamap/blob/37bd11cc4c57da931b7b117f829fb663d46ce480/rdma_map.go#L348-L368

gauravkghildiyal

Thanks @neaggarwMS. Mostly looks good.

(Please take a look at the presubmits to sign the CLA)

/ok-to-test

gauravkghildiyal · 2026-01-28T23:55:01Z

pkg/driver/rdmadevice.go

+
+		//Fallback to sysfs check if rdmamap fails. This is particularly related to a known
+		// issue to detect RDMA devices for certain Mellanox NICs
+		// https://github.com/Mellanox/rdmamap/issues/15
+
+		rdmaDir := filepath.Join("/sys/class/net", ifName, "device/infiniband")
+
+		entries, err := os.ReadDir(rdmaDir)
+		if err != nil {
+			return "", fmt.Errorf("no RDMA device for %s: %w", ifName, err)
+		}
+
+		for _, entry := range entries {
+			if entry.IsDir() {
+				return entry.Name(), nil // Return first RDMA device found (e.g., "mlx5_0")
+			}
+		}
+
+		return "", fmt.Errorf("no RDMA device found for %s", ifName)
+	}
+
+	return rdmaDev, nil


+1

Extending this further, can we please refactor the existing function

dranet/pkg/inventory/sysfs.go

Line 112 in 97d40e9

func hasRDMADeviceInSysfs(ifName string) bool {

into two functions, similar to the upstream implementation of IsRDmaDeviceForNetdevice and GetRdmaDeviceForNetdevice, where IsRDmaDeviceForNetdevice is a simple wrapper over GetRdmaDeviceForNetdevice (Ref. https://github.com/Mellanox/rdmamap/blob/37bd11cc4c57da931b7b117f829fb663d46ce480/rdma_map.go#L348-L368

gauravkghildiyal · 2026-02-04T22:04:09Z

Hey @neaggarwMS, checking in to see if you still need this change and will be able to accommodate the minor proposed comment.

aojea · 2026-02-05T14:00:31Z

ping @neaggarwMS , please check last comment and also sign the CLA, this seems almost ready to go

Thanks

Fix PrepareResourceClaim to read the rdmadev from sysfs path as well

f15af9c

k8s-ci-robot requested review from MikeZappa87 and danwinship January 28, 2026 00:23

k8s-ci-robot added the needs-ok-to-test Indicates a PR that requires an org member to verify it is safe to test. label Jan 28, 2026

k8s-ci-robot added cncf-cla: no Indicates the PR's author has not signed the CNCF CLA. size/L Denotes a PR that changes 100-499 lines, ignoring generated files. labels Jan 28, 2026

tamilmani1989 reviewed Jan 28, 2026

View reviewed changes

gauravkghildiyal reviewed Jan 28, 2026

View reviewed changes

k8s-ci-robot added ok-to-test Indicates a non-member PR verified by an org member that is safe to test. and removed needs-ok-to-test Indicates a PR that requires an org member to verify it is safe to test. labels Jan 28, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix PrepareResourceClaim to read the rdmadev from sysfs path as well#57

Fix PrepareResourceClaim to read the rdmadev from sysfs path as well#57
neaggarwMS wants to merge 1 commit intokubernetes-sigs:mainfrom
neaggarwMS:main

neaggarwMS commented Jan 28, 2026

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

linux-foundation-easycla bot commented Jan 28, 2026

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

tamilmani1989 Jan 28, 2026

Uh oh!

gauravkghildiyal Jan 28, 2026

Uh oh!

gauravkghildiyal left a comment

Uh oh!

gauravkghildiyal Jan 28, 2026

Uh oh!

gauravkghildiyal commented Feb 4, 2026

Uh oh!

aojea commented Feb 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Conversation

neaggarwMS commented Jan 28, 2026

Changes:

Testing:

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

linux-foundation-easycla bot commented Jan 28, 2026

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

k8s-ci-robot commented Jan 28, 2026

Uh oh!

tamilmani1989 Jan 28, 2026

Choose a reason for hiding this comment

Uh oh!

gauravkghildiyal Jan 28, 2026

Choose a reason for hiding this comment

Uh oh!

gauravkghildiyal left a comment

Choose a reason for hiding this comment

Uh oh!

gauravkghildiyal Jan 28, 2026

Choose a reason for hiding this comment

Uh oh!

gauravkghildiyal commented Feb 4, 2026

Uh oh!

aojea commented Feb 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants