support mig devices #53
Merged
Conversation
This file contains 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
…. Adding some enhanced error catching for nvml queries
Signed-off-by: Bram Vogelaar <bram@attachmentgenie.com>
Signed-off-by: Bram Vogelaar <bram@attachmentgenie.com>
|
spot check: |
schmichael
approved these changes
Aug 21, 2024
Comment on lines
+181
to
+188
| // A30/A100 MIG devices have no stats. | ||
| // | ||
| // https://docs.nvidia.com/datacenter/tesla/mig-user-guide/#telemetry | ||
| // | ||
| // Is this fixed on H100 or later? Maybe? | ||
| if mode == mig || mode == parent { | ||
| continue | ||
| } |
There was a problem hiding this comment.
Would it be safe to attempt to call DeviceInfoAndStatusByUUID and log/continue on error? I'd just hate for this to be something NVidia fixes in a driver update and then our plugin languishes for months without support because we don't even try.
There was a problem hiding this comment.
If we log we end up spamming a log line for each MIG device for each period. In the sad case that's 7 MIG devices for 8 GPUs every 30 seconds which is a lot of log spam for the hope Nvidia will fix their stuff.
nvml/driver_linux.go
Outdated
| } | ||
| utzEncU := uint(utzEnc) | ||
| memUsedU := mem.Used / (1 << 20) |
There was a problem hiding this comment.
Comment and/or a const for 1 << 20
Co-authored-by: Michael Schurter <michael.schurter@gmail.com>
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Outdated suggestions cannot be applied.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.
Incorporates previous PRs by @attachmentgenie and @isidentical while fixing a couple of bugs and adding MIG specific tests to the mock driver implementation.
Closes #3
Closes #27
Closes #40