Skip to content

add cached behavior to docs #1686

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

Open
alt-dima wants to merge 1 commit into
base: main
Choose a base branch
from
Open

add cached behavior to docs #1686

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
7 changes: 7 additions & 0 deletions content/docs/2.19/reference/scaledobject-spec.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -184,6 +184,13 @@ When using `behavior` with value `currentReplicasIfLower`, the current number of

**Example:** When my Prometheus instance becomes unavailable 3 times in a row, KEDA changes the HPA metric to scale the deployment to 3 replicas when I have `fallback.replicas` set to 6, but the current replicas are 3, with a `behavior` 'currentReplicasILower'.

#### `cached` behavior
When using `behavior` with value `cached`, KEDA will use the last known metric value from the cache. If cached metrics are available, those metrics will be returned instead of calculating fallback replicas. If no cached metrics are found, the value of `fallback.replicas` will be used as a fallback.

This behavior is most useful when your ScaledObject has multiple scalers. For example, if you have both a Prometheus scaler and a CPU scaler, and the Prometheus instance becomes unavailable, the `cached` behavior allows the HPA to continue scaling based on CPU metrics while using the last known Prometheus metric value. Without this behavior, if one scaler becomes unavailable, the HPA stops scaling entirely.

**Example:** When my Prometheus instance becomes unavailable 3 times in a row and I have `behavior` set to 'cached', KEDA will return the last known metric value from the cache. If the last cached metric would have caused the deployment to scale to 8 replicas, it will continue to target 8 replicas. If no cached metrics exist, it falls back to `fallback.replicas`.

## advanced

### restoreToOriginalReplicaCount
Expand Down