Chapter 4. Event-Driven Ansible

Event-Driven Ansible is a new way to enhance and expand automation by improving IT speed and agility while enabling consistency and resilience. Event-Driven Ansible is designed for simplicity and flexibility.

Known issues

  • Both contributor and editor roles cannot set the AWX token. Only users with administrator roles can set the AWX token.
  • Activation-job pods do not have request limits.
  • The onboarding wizard does not request a controller token creation.
  • Users cannot filter through a list of tokens under the Controller Token tab.
  • Only the users with administrator rights can set or change their passwords.
  • If there is a failure, an activation with restart policy set to Always is unable to restart the failed activation.
  • Disabling and enabling an activation causes the restart count to increase by one count. This behavior results in an incorrect restart count.
  • You must run Podman pods with memory limits.
  • Users can add multiple tokens even when only the first AWX token is used.
  • A race condition occurs when creating and rapidly deleting an activation causes errors.
  • When users filter any list, only the items that are on the list get filtered.
  • When ongoing activations start multiple jobs, a few jobs are not recorded in the audit logs.
  • When a job template fails, a few key attributes are missing in the event payload.
  • Restart policy in a Kubernetes deployment does not restart successful activations that are marked as failed.
  • An incorrect status is reported for activations that are disabled or enabled.
  • If the run_job_template action fails, the rule is not counted as executed.
  • RHEL 9.2 activations cannot connect to the host.
  • Restarting the Event-Driven Ansible server can cause activation states to become stale.
  • Bulk deletion of rulebook activation lists is not consistent, and the deletion can be either successful or unsuccessful.
  • When users access the detail screen of a rule audit, the related rulebook activation link is broken.
  • Long running activations with loads of events can cause an out of disk space issue. Resolved in installer release 2.4-6.
  • Certain characters, such as hyphen (-), forward slash (/), and period (.), are not supported in the event keys. Resolved in installer release 2.4-3.
  • When there are more activations than available workers, disabling the activations incorrectly shows them in running state. Resolved in installer release 2.4-3.
  • Event-Driven Ansible activation pods are running out of memory on RHEL 9. Resolved in installer release 2.4-3.
  • When all workers are busy with activation processes, other asynchronous tasks are not executed, such as importing projects. Resolved in installer release 2.4-3.