Resources
🤖 For AI agents: The complete documentation index is available at /docs/llms.txt. A markdown version of this page is available at /docs/cloud/reference/limits.md.

FusionAuth Cloud Limits

Edit on GitHub
View Markdown

FusionAuth Cloud has the same limitations as self-hosted FusionAuth.

As a managed service, FusionAuth Cloud is subject to the following additional limitations:

  • No access is provided to the server on which your deployment is running. This includes access to the database, Elasticsearch, or ssh. You can access your data via FusionAuth API or the admin UI. To install a password hashing plugin, please open a support ticket.
  • There is no API to manage FusionAuth Cloud deployments.
  • You cannot modify any of the FusionAuth configuration options.
  • You cannot self-service downgrade the version of a FusionAuth Cloud deployment. For example, you cannot change the version from 1.35.0 to 1.34.0.
  • You cannot run a Kickstart file on a FusionAuth Cloud deployment.
  • There is no support for proxy customization to add, for example, tenant routing. To accomplish such goals, add your own proxy layer such as CloudFlare, with FusionAuth Cloud as an origin. Make sure you configure the trusted proxies.
    • You cannot modify X-Forwarded-Port or X-Forwarded-Proto. For example, you can't proxy a FusionAuth Cloud instance to make it appear as if it was running at http://localhost or another non-TLS endpoint.
  • Use of port 25 is not allowed. To connect to an SMTP server such as Mailgun or SES, use a different port.
  • If you are on FusionAuth Cloud and you find that some requests are failing with a 429 status code, you are being rate limited. This isn't intentional, but an automated part of our infrastructure to ensure FusionAuth Cloud performance and security. Learn more here, including how to avoid rate limiting for certain servers.
  • If you want to run Advanced Threat Detection, an Enterprise feature, you'll need a cloud deployment with sufficient memory. It must be a Large or X-Large.
  • With HA and other multi-node deployments, requests are passed through a load balancer. When making requests to node specific metrics endpoints such as /api/status or /api/prometheus/metrics each request may return different results because the response is specific to the service node responding to the request. Using these endpoints to monitor a deployment is not recommended.
  • You may not modify the Elasticsearch settings or view the Elasticsearch index directly. Among other things, this means that you can't use some of the troubleshooting steps available to users self-hosting FusionAuth.
  • OpenTelemetry data is not available on FusionAuth Cloud deployments.
  • There is a limit of 1000 indexed fields. These include user.data, registration.data and standard indexed fields like email.
  • In certain cases, only current log files are available for download under System -> Logs . If you need all log files, including those previously rotated, please open a support ticket.