Resolved Issues¶

This section lists a number of resolved issues. For a complete list, see the Fuel for OpenStack Launchpad project.

• Added possibility to control restries count for Puppet-based tasks. See example. See LP1457794.

• Fixed the formula for the placement groups count in Ceph. See LP1464656.

• You can now create and manage vitual IP addresses through Nailgun and fuel client. See LP1482399.

• Fixed the issue with MySQL server going down because of the incorrect processing of the ignore-db-dir option. See LP1484552.

• Added an ability to connect to virtual machines in an OpenStack environment integrated with VMware vSphere through the VNC client. See LP1511422.

• Added support for the be2net-dkms kernel module on the HP ProLiant BL460c server on Ubuntu 14.04. See LP1533501.

• Fixed the rabbitmqctl issue that caused atom table overflow in Erlang VM. See LP1534519.

• Enabled the user to specify the disk on which to install the Fuel Master node and whether to format the disk before the installation or not by passing the installdrive and forceformat parameters to the kernel using the grub command line in the Fuel installation menu. See LP1535712.

• Enabled deployment of the OpenStack environments without controller nodes. Thereby, the user can replace controller roles with custom roles using Fuel plugins. See LP1538233.

• Fixed the issue for Swift/v1 endpoint containing an IP address instead of a FQDN in cases when Ceph was deployed with RadosGW. See LP1540133.

• Introduced a feature that allows specifying the order of network templates execution for a deployment. See LP1540374.

• Added the capability to forcefully update the settings and the attributes of an environment in the operational state by implementing an optional --force flag to the following commands:

  fuel settings --env <env_id> --upload --force
  fuel env --env <env_id> --attributes --upload --force
  

  See LP1540434.

  Also implemented the fuel redeploy-changes and fuel2 redeploy commands to apply the changes to an environment in the operational state.

  fuel redeploy-changes --env <env_id>
  fuel2 env redeploy <env_id>
  

  See LP1540558.

• Enabled compute-vmware and cinder-vmware roles combination for the same node. See LP1540756.

• Now, network_scheme contains information about the network gateway if it is defined in network-groups. See LP1549034.

• You can now allow removing the default gateway by overriding it as an empty string. See LP1549550.

• Nodes do not enter the error state when running puppet apply and detecting other Puppet instances. See LP1552805.

• Previously, during the upload of a new network YAML file, an error might appear pointing to the type of networks with the wrong configuration. Having many networks, it was difficult to determine which network had issues. Now, the error message contains network IDs. See LP1554106.

• Added the capability to override the primary-controller, controller, and compute node roles through Hiera for the openstack-network manifests. See LP1554796.

• Changed the Fuel serialization process to fix the Nailgun behavior in VLAN network environments with multiple node groups and network templates. Previously, it was impossible to deploy such a configuration due to a broken serialization module. See LP1556917.

• Previously, when specifying heartbeat and handshake_timeout parameters in rabbitmq.config, the default parameters disappeared. Now, rabbit.* ``configuration options can be handled by ``rabbit_config_variables in Hiera. See LP1560687.

• The order of tasks in the YAML file in /openstack-network/ is now strict to avoid incorrect functioning when introducting new elements. See LP1561648.

• Added an ability for plugin deployment tasks to depend on the settings of another plugin even if the latter is not installed. See LP1564131.