Release v0.28.0

v0.28.0

• Add a flag to enable verbose logs.
• Replace pods when changing node selectors.
• Add an option to automatically replace failed processes.
• Allow deleting pods when other pods are stuck in Terminating.
• Refactor the creation of service, pods, and PVCs to be based around the
  process group status.
• Enable leader election by default.
• Use the FDB client to read the JSON status, rather than the CLI.
• Automatically build and publish docker images for the operator.
• Coordinate bounces across data centers when doing upgrades.

Release v0.27.1

v0.27.1

• Prevent the operator from crashing on errors in lock acquisition.
• Allow deleting other pods when we have a pod stuck in a Terminating state.