Skip to content

Conversation

@klueska
Copy link
Collaborator

@klueska klueska commented Oct 11, 2024

This component is responsible for publishing ResourceSlices which advertise IMEX channels for each discovered IMEX domain in a cluster.

Copy link
Contributor

@guptaNswati guptaNswati left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. not sure, if we need a timeout condition to handle any lags in the creation and deletion of the new and old imex channels. I do see the 10 min time sync.

@klueska
Copy link
Collaborator Author

klueska commented Oct 12, 2024

The first node that pops up that is part of a unique IMEX domain, immediately triggers the creation of a ResourceSlice with IMEX channels for that domain. The last node that is part of a unique IMEX domain to disappear, immediately triggers the deletion of the ResourceSlice with IMEX channels for that domain.

The 10min time-sync is how often the informer forces a manual sync of its cache. Under normal operation every add/update/delete triggers an update of this cache, but this timed-sync ensures that it is also done periodically in the absence of these events (and to avoid potential races / bugs, of which there are no known ones).

@klueska klueska force-pushed the add-imex-controller branch 4 times, most recently from 437f309 to 56e6f85 Compare October 14, 2024 12:09
@klueska klueska force-pushed the add-imex-controller branch from 56e6f85 to d9cb0c5 Compare October 14, 2024 12:19
@klueska klueska force-pushed the add-imex-controller branch from d9cb0c5 to 53e24ab Compare October 14, 2024 12:28
@klueska klueska merged commit 7ddfa53 into NVIDIA:main Oct 15, 2024
6 checks passed
@klueska klueska added this to the v25.3.0 milestone Aug 13, 2025
@klueska klueska deleted the add-imex-controller branch August 20, 2025 21:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Backlog

Development

Successfully merging this pull request may close these issues.

2 participants