A few weeks ago we started distributing a brand new event-tracking and auditing system in Atlas. Today we are excited to announce that the new Atlas event and auditing system is available to all Atlas users. Read on to learn more about how we redesigned events to be easily consumable by operators and valuable for teams who want to understand infrastructure changes over time.
If you previously used the event logging in Atlas the following screenshot should seem familiar:
There were a number of challenges in implementing our first event system, and we have taken those challenges as learning opportunities in this new system.
One common feedback item about the previous event system was that it was very difficult to scan and parse as a human. We designed the previous event system to behave much like a speadsheet, but that turned out to be the wrong user experience.
In the new system, each event is a human-friendly format with the important components in bold and linked if possible. Below you can see the old event system (bottom) and the new event system (top) for a quick comparison.
In the future, we are going to extend the functionality to export to CSV and other machine-parsable formats, but we wanted the user interface for the new event system to be human-first. If you need CSV or other machine-parsable format of your event log, please contact support.
Improved Information Display
Another common feedback item was that the prior event system displayed too much information. For some users, this hid important events from view. Even moderate Atlas use would result in hundreds of events, making it very difficult to distinguish important events from traditional "logs".
One of the biggest changes in the new event system is that we carefully capture only the most important events to users and display them in the logs. As part of this transition, we are able to capture even more information (such as attribute diffs) that was not previously possible.
The new event system also intelligently tracks deleted resources. Even after you have deleted something, there is still an audit trail with its original name and metadata information stored as an event.
In addition to capturing more information about the user, browser, and IP address, the new event system calculates attribute diffs when possible. For example, if you update a variable, rename a resource, or update a description, you will see a diff or all the updated attributes:
This allows administrators to see and track changes, even after changes have been applied. All resource diffs are securely encrypted using Vault since attribute diffs could contain sensitive data.
The new event system is currently live for all users. To view the event system, you will need to have "admin" permissions on the resource and then visit the "Event log" from the resource page.
In addition to viewing your own personal event log, you can view all events in a particular organization provided you are an owner of that organization. Simply click on your organization in the account settings page and visit the "Event log" for that organization.
Lastly, each resource also has its own event log. While the organizational view of all events may be helpful, a more granular event log is available per-resource. If you have admin permissions for a resource, you can click the "Event log" item in the sidebar to view all events which correspond directly to that resource (such as an environment or build configuration).
We would like to extend a special thank you to all the customers who helped us beta test this new event system and provide valuable feedback. If you or your organization is interested in trying new beta features in Atlas, get in touch by emailing email@example.com.