ejabberd-0.6.0

0.6.0 - 2023-08-30

Added

• Extend CI to include cluster chaos, e.g. pod kills, failures, etc., processone
  rtb to test XMPP connections as well as scaling up and down.
• Use kubernetes leases to elect an ejabberd leading pod for clustering. This
  improves the robustness of the helm chart overall. The elector is enabled at
  default. If disabled, it falls back to the old k8s-DNS-based method.
• Add extra sections in values.yml to define host_config and
  append_host_config.

Changed

• BREAKING:
  Refactor .Values.modules to allow enabling/disabling, configuring and adding
  of ejabberd modules.
• Move image run scripts into the chart templates. This enables admins to easier
  exchange the container image.
• Bumb ejabberd image to 78f81de252dc932cd47b91d1a84ca8e8f0647498-k8s6.

Fixed

• Fix parsing host values.

Removed

• Remove the possibility to change the TCP healthcheck port. Healthchecks are
  now performed with a custom healthcheck script. You still can define custom
  healthchecks within the values.yml file.