Release 2.2.0

Bugfixes on top of 2.2.0rc1

• Do not store vm as peristant xml to prevent operations to start vms that have been started on another node
• Make it possible to list sizes based on the location code
• Capacityplanner now also considered reserved RAM for physical machine
• Change default libvirt machine type to ubuntu to support modern OS like Windows 2016
• Prevent portforward creation for vms that do not belong to CloudSpace which is passed in API
• Fix UI issue to switch selected CloudSpace when changing vm id in the url
• Live migration does not require CPU nodes hostnames to be resolvable and forces datapath over backplane1
• Stack Overview page shows summary of used memory, running vms, and routeroses
• Ensure cleanup of vdisks in case of VM creation failure

Release 2.2.0rc1

The main addition of release 2.2.0 compared to 2.1.7 is the added support of OpenvStorage Enterprise Edition. See docs for configuration of OpenvStorage.

Bugfixes

• Don't show jobs nobody is waiting for as timeout.
• Reorganize capacity planner to consider CPU nodes with more memory than others.
• Adjust timing policy of several healthchecks.
• Make portforwarding more performant when having 10+ per cloudspace.
• Hide destroyed objects by default (Accounts, Cloudspaces and Virtual Machines) in operation portal.
• Improve maintenance of storagenode.

Full changelog: see