fix(ascend): mount the device NIC addresses so RoCE endpoints can be built - #28
Merged
Merged
Conversation
…built A transport building Device RoCE endpoints reads each NPU's NIC address from /etc/hccn.conf, or from hccn_tool, which lives in a driver directory none of the injected mounts covers. Neither reaches the container, so the engine fails to initialize with "Failed to get device ip from hccn.conf and hccn_tool" and every process that opens the transport dies at startup: a shared KV cache store on an Ascend node never comes up, and an engine attaching to one fails the same way. Inject the file beside the ranktable it already sits next to. A host that configured no device NICs has no such file and path_to_cdi_mount skips what is absent, so a container that never opens the transport sees no change.
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request adds the host configuration file path /etc/hccn.conf (_HCCN_CONF_PATH) and includes it in the mount paths for the Ascend CDI deployer. This ensures containers using Device RoCE can access host NIC addresses and initialize successfully. There are no review comments, so no additional feedback is provided.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A transport building Device RoCE endpoints reads each NPU's NIC address from /etc/hccn.conf, or from hccn_tool, which lives in a driver directory none of the injected mounts covers. Neither reaches the container, so the engine fails to initialize with "Failed to get device ip from hccn.conf and hccn_tool" and every process that opens the transport dies at startup: a shared KV cache store on an Ascend node never comes up, and an engine attaching to one fails the same way.
Inject the file beside the ranktable it already sits next to. A host that configured no device NICs has no such file and path_to_cdi_mount skips what is absent, so a container that never opens the transport sees no change.
Ref:
https://gitcode.com/cann/hixl/wiki/Mooncake%EF%BC%88NPU%20%E7%89%88%EF%BC%89%E5%AE%8C%E6%95%B4%E6%8C%87%E5%8D%97.md#23-%E6%B3%A8%E6%84%8F%E4%BA%8B%E9%A1%B9