3scale - What happens if the network between apicast and api manager is lost?

Solution Unverified - Updated -

Environment

  • Red Hat 3scale API Management 2.11

Issue

  • If api manager and api gateway are unable to communicate due to a network issue or other outage:
    • Will the api gateway still work?
    • Will there be any effect on the service such as whether monitoring information will be lost or not, can update configuration return?
    • If api gateway and api manager are not connected at some point due to network loss, how long will it continue to work or will it reconnect itself?

Resolution

  • Assuming default settings and that only the system containers are unreachable; APICast should continue working for up to 5 minutes, as this is how often APICast refreshes its configuration.
    It is only "up to 5 min" rather than "5 min" as how long you have depends on how long ago APICast last pulled its configuration down.
  • If the Backend pods are also down/unreachable then APICast will be unable to validate authentication data as well as being unable to send through metrics, this will cause your customers to receive authentication errors unless you have Authentication Caching or Batching Policy which provide a cache of authenticated credentials for APICast to compare to, updated regularly.
  • Note that the above is only for the production APICast instance. By default the Staging APICast instance pulls the configuration down for every call so if the System pods are unreachable it will fail immediately.

This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.

Comments