Platform9 3.11 release notes
Platform9 Managed Kubernetes
1. CoreDNS Support
As part of moving towards v1.13 of Kubernetes, Platform9 will now be supporting CoreDNS for use by default in 3.11 clusters. For clusters deployed pre-3.11, upgrading to 3.11 will automatically replace KubeDNS with CoreDNS.
2. UI-Based RBAC for K8s
Users will now be able to take advantage of Kubernetes RBAC functionality in the UI. This will allow users to create Roles, RoleBindings, ClusterRoles, RoleBindings, and apply granular operations and permissions to various users and groups across multiple clusters and namespaces. This is done without needing to do any YAML configurations in the CLI and is designed for scaled permissions across multiple environments simultaneously.
3. Calico General Availability
Calico backend is now GA. Note: Upgrading existing Calico clusters to 3.11 is not supported.
4. Canal (BETA) CNI Support
Platform9 has added support for Canal CNI plugin as a backend during cluster configuration. Managed Kubernetes now supports Calico, Flannel, and Canal
5. Custom AMI Support
Users will be able to specify their own AMI to be used during cluster provisioning for the Kubernetes cluster. Note: CentOS 7 and Ubuntu 16.04 are the only supported operating systems
6. Bug Fixes and Improvements
This release contains several performance optimizations and bug fixes, which should result in a better user experience for your Platform9 cloud platform. Some significant ones are listed below.
etcdv3 migration: Clusters on 3.11 will use etcdv3 as the primary datastore for Kubernetes. Existing clusters will continue to use etcdv2 as the primary datastore for Kubernetes upon upgrade
Update Docker to 18.09
Bump CNI Plugins version to 0.7.5
Two metrics-server deployments running after upgrade (fix this by removing the older on upgrade)
(fixed) Yum Package Conflict: ‘docker-ce-cli’. docker-ce-cli yum package conflict encountered if the customer runs a yum update on a PMK node
Tag AWS cluster subnets to guide k8s-created ELBs to correct one
qbert auditing support for log file creation on cluster-lifecycle based user logs
Platform9 Managed OpenStack
1. Event & Alerting Logging for Auditing
New in 3.11 is the ability to enable logging for events and alerts for:
- Nova API
- Glance API (Platform9 Control Plane and image library host)
- Neutron Server
- Heat API
- To enable and access each log file please contact your Platform9 Customer Success representative
- Self-Service access to event and alerting logs will come in a future release
2. SR-IOV Support via Openstack APIs
Network sensitive workloads and Virtualized Network Functions (VNF) can now be deployed and leverage SR-IOV compatible physical network interface cards for faster performance and lower latency.
3. Gravatar Profiles
Users that have configured a Gravatar image will now see their image within their Platform9 profile
4. vSphere 6.7 Support
Platform9 has tested and certified vSphere version 6.7
5. Network Enhancement: Open vSwitch High Availability using VRRP Support
Users can now configure layer-3 high availability to support active-active routers
6. Bug Fixes and Improvements
A Gnocchi memory consumption bug has been resolved.
A discovery issue with VMs and their metadata in vSphere has been resolved.
A Nova bug impacting the number of VMs discovered in vSphere has been resolved.
A bug causing volumes with pending tasks erroneously showing up in UI under a different tenant’s view has been fixed.
A bug impacting PF9-Discovery failing to establish connection to vCenter server has been resolved.
A UI performance issue when deleting multiple instances simultaneously has been addressed.
The host de-authorizing process has been improved to ensure all Platform9 artifacts are removed.
An issue preventing instances from booting after a volume is migrated has been resolved.
An issue causing the inability to resize instances from UI after deleting the existing flavor has been resolved.
An issue impacting port availability for reallocation after deleting a VM from vCenter has been resolved.
The vCenter Management Appliance now supports user configured proxy authentication.
The performance of the SSO Group management UI has been improved.
Failed snapshot attempts will now show up in UI under that Instance’s Events page, which will list all the actions performed on that instance along with their outcome.
Administrators can now see all instances on a host across all of their managed tenants under the Infrastructure Hosts tab.
Better error reporting in UI when a Murano application fails to deploy.
June 03, 2019