Bug Fixes

subnet configuration as list caused some issues #424
fixed master being used for compute despite useMasterAsCompute: False #455
host file bug fixed via yaml lock #464
made waitForCloudServices a cloud specific key to accommodate hybrid clouds #465 where services are not present on every cloud
if a worker with the same name as an already running server should be started by on demand scheduling, the existing server is terminated because for some reason it has decoupled from slurm. "Why" still needs to be investigated. Issue contained in #461
DNS not working on many clouds #509
NFSShare issues for certain folder structures due to an inspecific regex #507
fixed autoMount not working properly after update

Features

the timeout for waiting for the ssh availability of instance after its start is now a key sshTimeout in the configuration file #483 we also have added a key for later on demand scheduling is called cloudScheduling: sshTimeout: 42.
added dependabot and added other git project features
It is now also possible to start permanent workers that are not on demand using onDemand: False #487
It is now possible to use your own ansible.cfg without needing to replace it everytime a bibigrid update comes around. #485
It is now possible to use your own slurm.conf without needing to replace it everytime a bibigrid update comes around. #475
Slurm upgrade. The new method works by pinning the current version. That way on demand scheduled workers stay on the same version the whole BiBiGrid cluster does. However, this will be overwritten if you update the cluster via bibigrids update command #492
Added structures and changed ansible.cfg so that new ansible roles can be included easily making use of the userRoles key #439 (a dependency for #440 )
autoMount is now more advanced #506