Red Hat has updated its Ansible Automation Platform, pushing infrastructure management tool Ansible Tower over the 3.7-line and making good on an earlier promise to add an easier way to manage resources.

The latter was realised by putting the automation services catalog Red Hat previewed in April onto the platform. Similar to the service catalog some users might know from another Red Hat product, OpenShift, it allows admins to set up a sort of self-service portal for automation resources. 

The catalog can be populated with “products” like job templates, for example to provision virtual machines or request OpenShift projects, and workflows, which can be grouped into portfolios for different user groups. If a product has been requested by a user, set approvers get a notification and can then either give their thumbs up before granting the enquirer access, put the request on hold until availability is checked, or deny it altogether. 

Depending on the kind of resource, admins are free to automatically rope in more than one approver as some might be associated with cost and therefore might, for instance, need an okay from finance.

Ansible Tower, which is part of the automation platform, has also received some changes, though they are more about the user experience and therefore not as visible. To begin with, there’s now a “supported Ansible Tower collection” to facilitate the configuration of resources in Tower via playbooks. 

Under the hood Ansible Tower 3.7 has been reworked to allow project and inventory updates to happen while other jobs are running, which is meant to lead to less blockage. The Tower team also optimised how API queries and logs are processed, simplified the tool’s clustering architecture and replaced the external logging infrastructure to smooth out some UX kinks.

Other new additions include an automation calculator and a notification panel, both of which can be found under Automation Analytics in the left hand menu. The automation calculator is supposed to give users a way of putting a number on how much time and money automation saved them, which could always help to get more buy-in from superiors. 

And while the notification panel is a neat addition, the new notifications that go along with it seem to be a tad more interesting. They finally let users know when a node in a cluster is down – previously that could only be done for whole clusters – inform about jobs in error state, and nudge users when the job queue gets too long. 

More information about the update can be found in a number of blog posts via the general announcement.