Astronomer Outage History

We are currently investigating this issue.

We are currently observing elevated task failures and latency for some deployments running in the AWS us-east-1 region. This is related to an ongoing AWS incident affecting a single Availability Zone (use1-az4), where EC2 and EBS resources have experienced impairments. What to expect: You may see intermittent task failures or retries in your deployments. In most cases, these failures are transient and should self-resolve automatically as AWS continues recovery. What you should do: No immediate action is required. However, if you notice consistent or prolonged failures, please reach out to Astronomer Support, and we’ll help investigate further. We are continuing to monitor the situation closely and will share updates as needed. Refer: https://health.aws.amazon.com/health/status

We are investigating delayed delivery for some time-based alerts. The delay is primarily visible for DAG Timeliness alerts, and may also affect Observe-related SLA, Proactive SLA, and Data Quality monitor alerts. DAG Duration and Task Duration alerts do not appear to be affected at this time.

Attempting to access the dags page in the Airflow UI results in a 403 Forbidden error. This should not be affecting task execution.

We’re investigating an issue affecting Dashboard cost breakdown data. For affected customers, cost breakdown information may appear stale and may not have updated since May 1, 2026. Our team is actively investigating the cause and working to restore current data. We’ll share another update as we have more information.

We are continuing to investigate this issue.

It seems to be only in AWS clusters for now. We have a workaround we can apply while we investigate.

We are continuing to investigate this issue.

We have identified an issue with Astro Executor deployments where terminating workers can continue to accept new tasks, which can, in some cases, can cause these tasks to fail.

We are currently investigating an issue causing false positive alerts via Astro Alerts. Our team is actively investigating the issue.

We are currently investigating an issue with the Astronomer Cloud UI and API. User impact: The interface may be intermittently available at this time, and Airflow UIs appear to be inaccessible due to the inability to authenticate users through the API. Given the API's role in Deployment and Cluster management, those components are likely to be unresponsive during this time. Please stand by for updates. If you have any immediate questions, please reach out in Slack or open a ticket at support.astronomer.io.

The issue with our image repository is still being investigated, but most customers should be able to resume normal operations. Please subscribe for further updates.

We are continuing to investigate this issue.

Our image repository is currently unavailable. We're investigating the issue and will update as we know more.

Fix has been applied for docker-registry auth proxy.

We are investigating an issue with API Keys / Access Tokens. Impact: CI/CD pipelines will likely fail Workaround: Deploy using the astro CLI tool, if possible

We have observed long and unstable logins to the Cloud UI. Additionally, we see that deployments of new images are not working at least some of the time. We believe both of these to be related to the same root cause and are investigating.

We are currently investigating an issue where all deployments are being shown UnHealthy even if they are not. Please go ahead with opening airflow web UI even though it says "Airflow Unavailable" Any of existing deployments or updates are NOT affected. Only the status info is affected.

We are currently investigating an issue where some customers are not able to access https://cloud.astronomer.io. Pipeline and task executions are not affected.

The UI for cloud.astronomer.io is currently inaccessible and updates to deployments are currently not working. Both Airflow UI's and the Astronomer UI cannot be reached currently. Existing Airflow deployments will continue to operate, but they are not viewable in the browser.

We are continuing to investigate this issue.

We are currently investigating the issue.

Airflow and Astronomer UI are now available again and new deployments will function correctly now. Airflow deployments may show as unhealthy and you may see "Airflow Not Available" but this message is inaccurate. You can still click on the link to access your Airflow deployments. We are in the process of correcting this spurious "Airflow Not Available" warning.

We are currently investigating the issue. For most customers, this issue should not be noticeable. For some deployment patterns, this issue may cause the Kubernetes autoscaler to scale down without respect for any currently running tasks which would kill those tasks and render them as zombies. As a temporary workaround, each time you trigger a deployment using `astro deploy`, please also update your Worker Queue(s) from the Astro UI by selecting your Workspace, then selecting your Deployment, then clicking "Edit" on your Default Worker Queue, then clicking the "Update Queue" button.