Bug was introduced trying to solved the following issue:
> When scenarios where two HPAs reference the same object but have
different metric calculation (e.g. ingresses with different weights),
kube-metrics-adapter calculates the two metrics but always overrides
one of the metrics when saving it to the store.
This commit fixes a issue where without the added `return` the system
was continuing in an invalid states where `labels2metric` wasn't
properly initialized, causing the system to panic.
Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>
Having the test filename prefix exactly as the tested file name helps systems
to identify the correlation.
Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>
The metrics store, both the custom and external one, make heavy usage of
maps of strings to strings. This leads to a presumable confusing or at
least hard to read codebase.
This commit tries to increase readability and maintainability of the
metric stores by using custom types. This commit has no effect in
performance or functionality, it's pure aesthetics.
Co-authored-by: Jonathan Juares Beber <jonathanbeber@gmail.com>
Signed-off-by: Katyanna Moura <amelie.kn@gmail.com>
The current implementation of the metrics store for custom metrics uses
just the object name as key. When scenarios where two HPAs reference the
same object but have different metric calculation (e.g. ingresses with
different weights), kube-metrics-adapter calculates the two metrics but
always overrides one of the metrics when saving it to the store.
This commit implements the use of a labels hash in the custom metrics
store. This way, metrics are identified not just by the referenced
object, but also by its selectors.
Co-authored-by: Katyanna Moura <amelie.kn@gmail.com>
Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>
This commit adds two new collectors to the adapter:
- ClusterScalingScheduleCollector; and
- ScalingScheduleCollector
Also, it introduces the required collectors plugins, initialization
logic in the server startup, documentation and deployment example
(including the helm chart). A new config flag is created,
`-scaling-schedule`, and allows to enable and to disable the collection
of such metrics. It's disabled by default.
This collectors are the required logic to utilise the CRDs introduced in
the #284 pull request. It makes use of the kubernetes go-client
implementations of a [Store][0] and [Reflector][1].
[0]: https://pkg.go.dev/k8s.io/client-go/tools/cache#Store
[1]: https://pkg.go.dev/k8s.io/client-go/tools/cache#Reflector
Signed-off-by: Jonathan Juares Beber <jonathanbeber@gmail.com>
* This commit adds a --disregard-incompatible-hpas that makes the HPA
provider stop erroring out when a collector cannot be created for a
metric in a HPA. Useful when kube-metrics-adapter runs alongside another
metrics provider. Fixes issue #94.
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Make tests pass
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Wraps the Plugin Not Found error in a new type that can be checked by the caller of a function to determine if its contents should be logged or added as an event to the HPA, when this HPA is incompatible.
The disregardIncompatibleHPAs is now targetting only the log or addition of the same event.
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Invert if expression to select when we should log
CreateNewMetricsCollector errors: don't log when both conditions are true - it's not a PluginNotFoundError
and disregardIncompatibleHPAs flag is set to true. This way, if an error
is NOT PluginNotFoundError it will always be logged, and when it IS
PluginNotFoundError it will only be logged when
disregardIncompatibleHPAs is false.
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Remove redundant "whether to"
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Add test case for updating HPAs via HPA Provider while disregarding
incompatible HPAs.
Signed-off-by: Tomás Pinho <me@tomaspinho.com>
* Prevent leaking collectors when HPA gets updated
This fixes an issue where collectors would be leaking when HPAs are
getting updated.
Fix this by stopping the collector started for the previous version of
the HPA.
Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>
* Add tests to verify old collector is removed
Signed-off-by: Mikkel Oscar Lyderik Larsen <mikkel.larsen@zalando.de>
Adding tests for error cases in metric_store.go. This will primarily
protect against regression.
Signed-off-by: Muhammad Muaaz Saleem <muhammad.muaaz.saleem@zalando.de>
Making metrics expiration TTL configurable globally configurable.
This is so RemoveExpired can be tested. Not making the TTL configurable
per metric yet because it could change behaviour in production.
Signed-off-by: Muhammad Muaaz Saleem <muhammad.muaaz.saleem@zalando.de>
Testing that metrics get removed once expired. Tests are failing
because metric.TTL isn't configurable at the moment
Signed-off-by: Muhammad Muaaz Saleem <muhammad.muaaz.saleem@zalando.de>
Adding test cases to insert/list/get a namespaced and a non-namespaced
resource metric
Signed-off-by: Muhammad Muaaz Saleem <muhammad.muaaz.saleem@zalando.de>
* feat: expose basic go metrics via prometheus
Signed-off-by: Martin Linkhorst <martin.linkhorst@zalando.de>
* feat: expose basic counter of successful and failed collections and updates
Signed-off-by: Martin Linkhorst <martin.linkhorst@zalando.de>
