Feb 28, 17:21 UTC
Resolved - On 28th Feb between 12:15 UTC and 15:20 UTC Connect apps were unable to perform user-impersonation operations and failed to execute. This had a widespread impact on all product extensions based on the Connect platform resulting in failure when performing actions on behalf of a user.
We have identified the root cause of the fault and have resolved it by scaling up the infrastructure. The incident is now resolved.
Automation rules that triggered during the outage may have failed to run. Those rules will not be able to be replayed by customers or Atlassian. Our support engineers have details regarding which customers and automation rules that were impacted. However, support engineers will not be able to replay the failed automation rules.
If you have questions or require support - please reach out at https://support.atlassian.com/contact/.
We apologise to all of our developers, customers and users who were impacted during this time.
We will release a public incident review (PIR) here in the upcoming weeks. We will detail all that we can about what caused the issue, and what we are doing to prevent it from happening again.
Feb 28, 15:57 UTC
Identified - Starting from 13:15 UTC, all user impersonation operations done by Connect apps in all regions using ACT_AS_USER scope were failing.
We have scaled up the underlying service responsible for tokens generation and we're seeing an improvement in response times and success rates. We continue to investigate the root cause and will provide the next update by 18:00 UTC, in 2h from now.
Feb 28, 13:39 UTC
Investigating - We have identified a problem with Connect apps where user impersonation is not working properly. We are investigating the root cause and will provide another update by 14:30 UTC.