Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Split scheduler metrics by resource #3425

Merged
merged 9 commits into from
Mar 6, 2024
Merged

Split scheduler metrics by resource #3425

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
8ece055
Split scheduler metrics by resource
Mo-Fatah Feb 27, 2024
90da54c
tidy up
Mo-Fatah Feb 27, 2024
9506d1f
Add resourceRenaming config
Mo-Fatah Feb 28, 2024
bba578e
lint
Mo-Fatah Feb 28, 2024
6daa0f3
Create a map of tracked resources metrics
Mo-Fatah Mar 5, 2024
6fc3ecf
Add metric name validation regex
Mo-Fatah Mar 5, 2024
99dc18d
lint
Mo-Fatah Mar 5, 2024
316a5c6
add jobs metrics to the counter map
Mo-Fatah Mar 5, 2024
923e81a
Merge branch 'master' into split-metrics-by-resource
Mo-Fatah Mar 5, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions config/scheduler/config.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,10 @@ schedulerMetrics:
- "memory"
- "ephemeral-storage"
- "nvidia.com/gpu"
resourceRenaming:
nvidia.com/gpu: "gpu"
amd.com/gpu: "gpu"
ephemeral-storage: "ephemeralStorage"
matchedRegexIndexByErrorMessageCacheSize: 100
resetInterval: "1h"
pulsar:
Expand Down
6 changes: 6 additions & 0 deletions internal/scheduler/configuration/configuration.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,6 +73,12 @@ type MetricsConfig struct {
TrackedErrorRegexes []string
// Metrics are exported for these resources.
TrackedResourceNames []v1.ResourceName
// Optionally rename resources in exported metrics.
// E.g., if ResourceRenaming["nvidia.com/gpu"] = "gpu", then metrics for resource "nvidia.com/gpu" use resource name "gpu" instead.
// This can be used to avoid illegal Prometheus metric names (e.g., for "nvidia.com/gpu" as "/" is not allowed).
// Allowed characters in resource names are [a-zA-Z_:][a-zA-Z0-9_:]*
// It can also be used to track multiple resources within the same metric, e.g., "nvidia.com/gpu" and "amd.com/gpu".
ResourceRenaming map[v1.ResourceName]string
Copy link
Contributor

Choose a reason for hiding this comment

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

We're using the validator package for config validation elsewhere. https://pkg.go.dev/github.com/go-playground/validator/v10#section-readme
Could you use this pre-defined regex to validate that these metric names are valid? https://github.com/go-playground/validator/blob/v10.19.0/regexes.go#L7

Copy link
Collaborator Author

@Mo-Fatah Mo-Fatah Mar 5, 2024 •
edited
Loading

Choose a reason for hiding this comment

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

this doesn't include underscores or colons. Also, it allows a resource that starts with a digit, which prometheus will reject later in the metric name.

most tracked resources don't include underscores or colons themselves, however this might reject valid resource names or accept invalid ones in the future.

The regex i added acts the same way as prometheus naming check (in the current version we are using ) https://github.com/prometheus/common/blob/49b36038aead362ce78f09337dc5872000bd048a/model/metric.go#L97

Copy link
Collaborator Author

@Mo-Fatah Mo-Fatah Mar 5, 2024 •
edited
Loading

Choose a reason for hiding this comment

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

i can use that function IsValidMetricName if you want to delegate the check to prometheus

// Controls the cycle time metrics.
// TODO(albin): Not used yet.
CycleTimeConfig PrometheusSummaryConfig
Expand Down
Loading
Loading