Initialize NVML on demand. #1969
Merged
Conversation
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
Contributor
Author
dashpole
reviewed
Jun 18, 2018
accelerators/nvidia.go
Outdated
| if !nm.nvmlInitialized { | ||
| initializeNVML(nm) | ||
| nm.Unlock() | ||
| } else { |
Collaborator
There was a problem hiding this comment.
you dont need the else here. Just unlock outside the if statement.
Contributor
Author
There was a problem hiding this comment.
Actually this made me realize that I can simplify the taking locks multiple times to taking lock just once. PTAL.
dashpole
reviewed
Jun 18, 2018
Collaborator
dashpole
left a comment
dashpole
left a comment
There was a problem hiding this comment.
one nit. Otherwise looks good.
Contributor
|
/lgtm |
rohitagarwal003
force-pushed
the
fix-race
branch
from
4c45a2e to
844caff
Compare
Earlier if the NVIDIA driver was not installed when cAdvisor was started we would start a goroutine to try to initialize NVML every minute. This resulted in a race. We can have a situation where: - goroutine tries to initialize NVML but fails. So, it sleeps for a minute. - the driver is installed. - a container that uses NVIDIA devices is started. This container would not get GPU stats because a minute has not passed since the last failed initialization attempt and so NVML is not initialized.
rohitagarwal003
force-pushed
the
fix-race
branch
from
844caff to
2ce4161
Compare
rohitagarwal003
added a commit
to rohitagarwal003/kubernetes
that referenced
this pull request
Jun 22, 2018
The race condition that required this sleep was fixed in google/cadvisor#1969. That was vendored in kubernetes#65334.
k8s-github-robot
pushed a commit
to kubernetes/kubernetes
that referenced
this pull request
Jun 22, 2018
Automatic merge from submit-queue (batch tested with PRs 65377, 63837, 65370, 65294, 65376). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>. Remove unneeded sleep from test. The race condition that required this sleep was fixed in google/cadvisor#1969. That was vendored in #65334. ```release-note NONE ``` /assign @jiayingz @vishh
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Earlier if the NVIDIA driver was not installed when cAdvisor was started
we would start a goroutine to try to initialize NVML every minute.
This resulted in a race. We can have a situation where:
This container would not get GPU stats because a minute has not passed
since the last failed initialization attempt and so NVML is not
initialized.