The operator currently uses the following default values for leader election:

• lease duration: 15s
• renew deadline: 10s
• retry period: 2s

In clusters with a very slow API server, these defaults are too short, causing the operator to restart because it loses the leader lease (see OSSM-6397).

Although the correct solution is to figure out why the API server is slow, we should allow admins to increase the default values until the problems with the API server are resolved.

We should introduce the following command-line options to the operator:

• --leader-elect-lease-duration=15s
• --leader-elect-renew-deadline=10s
• --leader-elect-retry-period=2s

Users who want to set these options, should modify the operator's ClusterServiceVersion.