KEP-5304: Promote DRA Discoverable Device Metadata API to Beta#6160
KEP-5304: Promote DRA Discoverable Device Metadata API to Beta#6160alaypatel07 wants to merge 2 commits into
Conversation
8dfd39a to
1237604
Compare
| - Metric name: `dra_metadata_feature_enabled` | ||
| - Description: Gauge (0/1) indicating if the feature is enabled on this driver instance | ||
| - Aggregation method: current value per node/driver | ||
| - Components exposing the metric: DRA driver plugin framework |
There was a problem hiding this comment.
... which may or may not be exposed by a DRA driver using the framework. Admins need to check the documentation of a DRA driver to determine how to collect these metrics.
That all of these metrics are by DRA driver is not clear from the KEP. The usual expection is that these are metrics in a Kubernetes component, which isn't true here.
There was a problem hiding this comment.
I agree, I have added an explicit section few lines about to reflect this.
| These goals will help you determine what you need to measure (SLIs) in the next | ||
| question. | ||
| --> | ||
| This feature adds minimal overhead to pod startup. SLOs expressed as metrics queries: |
There was a problem hiding this comment.
Devil's advocate: what the SLI is covering is only the overhead added to a DRA driver. It does not cover the overhead of doing more work for mounting files (container runtime, kernel).
That additional overhead is harder to measure. My expectation is that it is higher than merely writing some files.
It would be nice if the KEP at least acknowledged that this other overhead exists.
There was a problem hiding this comment.
+1 good point, acknowledged it in the KEP
1237604 to
c6809f5
Compare
Assisted-by: Cursor <no-reply@cursor.com> Signed-off-by: Alay Patel <alayp@nvidia.com>
c6809f5 to
a8c148b
Compare
| extending the production code to implement this enhancement. | ||
| --> | ||
|
|
||
| - `<package>`: `<date>` - `<test coverage>` |
There was a problem hiding this comment.
Can you fill this as part of the beta reqs?
sohankunkerkar
left a comment
There was a problem hiding this comment.
I have two comments but the PRR questionnaire is thoroughly filled in for Beta.
LGTM from the PRR shadow side.
Co-authored-by: Wendy Ha <139814343+wendy-ha18@users.noreply.github.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: alaypatel07, mrunalp The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
Assisted-by: Cursor no-reply@cursor.com