1. Proposed title of this feature request

Allow tuning on-prem keepalived check script period

2. What is the nature and description of the request?

Currently, the check script for the on-prem keepalived static pod runs the curl checks with a hard-coded 1 second frequency and a timeout of 0.9 seconds (so that it fits within the 1 second period).

We are requesting to make this period tunable so that, for example, one can set a 2 or 5 second period for the curl script and the curl timeout would be automatically adjusted to 100ms less (e.g. if period is 2, curl timeout is 1.9, if period is 5, curl timeout is 4.9...).

3. Why does the customer need this? (List the business requirements here)

Having a fixed curl period and timeout for the keepalived check script can be harmful in some heavily loaded scenarios, not only because of direct overload but because the check script can get stuck in really non-obvious ways. Being able to customize this may allow for interesting mitigation on those scenarios.

4. List any affected packages or components.

Keepalived