katanemo / arch

Arch is an intelligent prompt gateway. Engineered with (fast) LLMs for the secure handling, robust observability, and seamless integration of prompts with APIs - all outside business logic. Built by the core contributors of Envoy proxy, on Envoy.
https://archgw.com
Apache License 2.0
367 stars 19 forks source link

Developer app server errors are not sent back correctly #176

Closed adilhafeez closed 4 days ago

adilhafeez commented 1 week ago

Right now when a developer service fails due to some error (4xx or 5xx) envoy keeps on retrying and eventually sends back a 503. Which isn't a great experience. In case of error filter should short-circuit and sends error back.