Addressed issues

The following issues have been addressed in the MOSK 23.3 release:

  • [OpenStack] [34897] Resolved the issue that caused the unavailability of machines from the nodes with DPDK after update of OpenStack from Victoria to Wallaby.

  • [OpenStack] [34411] Resolved the issue with an incorrect port value for RabbitMQ after update.

  • [OpenStack] [25124] Improved performance while sending data between instances affected by the Multiprotocol Label Switching over Generic Routing Encapsulation (MPLSoGRE) throughput limitation.

  • [TF] [30738] Fixed the issue that caused the tf-vrouter-agent readiness probe failure (No Configuration for self).

  • [Update] [35111] Resolved the issue that caused the openstack-operator-ensure-resources job getting stuck in CrashLoopBackOff.

  • [WireGuard] [35147] Resolved the issue that prevented the WireGuard interface from having the IPv4 address assigned.

  • [Bare metal] [34342] Resolved the issue that caused a failure of the etcd pods due to the simultaneous deployment of several pods on a single node. To ensure that etcd pods are always placed on different nodes, MOSK now deploys etcd with the requiredDuringSchedulingIgnoredDuringExecution policy.

  • [StackLight] [35738] Resolved the issue with ucp-node-exporter. It was unable to bind port 9100, causing the ucp-node-exporter start failure. This issue was due to a conflict with the StackLight node-exporter, which was also binding the same port.

    The resolution of the issue involves an automatic change of the port for the StackLight node-exporter from 9100 to 19100. No manual port update is required.

    If your cluster uses a firewall, add an additional firewall rule that grants the same permissions to port 19100 as those currently assigned to port 9100 on all cluster nodes.