Skip to content
Open
Changes from all commits
Commits
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
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,13 @@ Tags are displayed on the delegate details page in Harness Manager. Go to [Tags

### DYNAMIC_REQUEST_HANDLING

If enabled, delegate will stop acquiring new tasks if the cpu/memory usage goes beyond values defined in `CPU_USAGE_THRESHOLD` and `MEMORY_USAGE_THRESHOLD`. The default value for both is 80.
Dynamic task handling is designed to prevent delegates from being overloaded. When enabled, a delegate will stop picking up new tasks if CPU (`CPU_USAGE_THRESHOLD`) or memory usage goes beyond the defined threshold. By default, both thresholds are set at 80%, and you can override either of them if needed.

If you change one value, the other continues to use the default. Even when dynamic task handling is not enabled, delegates may still reject tasks once the memory threshold is reached, which is expected behavior to keep the delegate stable.

:::note
The system periodically recomputes the list of tasks and then shares the updated list with delegates. This ensures that task distribution stays fair and efficient. Re-broadcasting of the recomputed list is supported, but it is available only through a feature flag. If you’d like this capability, please reach out to [Harness Support](mailto:support@harness.io).
:::

```yaml
- name: DYNAMIC_REQUEST_HANDLING
Expand Down