Is there an existing issue for this?

• I have searched the existing issues

Kong version ($ kong version)

2.5.1.0

Current Behavior

When calling kong.response.exit(), the function automatically sets the Content-Length header on the response even if the Transfer-Encoding header is present. This is in violation of RFC 7230 3.3.2:

A sender MUST NOT send a Content-

Improved or typo fixed.

https://github.com/apache/apisix/blob/0d4f65a9ae06430388af609ca0dcad386225d616/apisix/core/vault.lua

Solution

not doc for how to use vault with Apache APISIX

On contexts with many APIs and APIs calling other APIs in chains, having too many spans in the zipkin/jaeger dashboards may lead to a too complex usage of such dashboards and of not necessary big data volumes.

For Tyk's admin team, the steps representing the time spent in each Tyk middleware are much valuable.

But the upstream APIs teams just want to know the global time spent in Tyk interna

Version

1.10.x (beta)

Is your feature request related to a problem? Please describe.

In some cases, the gloo-fed deployment should be scaled down to 0, using helm chart values. For example, for disaster recovery, all Gloo Federation CRs should be restored before gloo-fed is scaled to 1, to ensure that gloo-fed doesn't delete placements of federated resources which were not yet restored.

See https://github.com/apiman/apiman/blob/master/.github/workflows/verify.yaml

Might as well also add builds for Java 11 and 15 while at it.

Follow up on #1245

Add a check to validate that x-extensible-enum properties have a string type.

Proposal

Background

There are two different sets of backend E2E code in the current project, each using a different implementation, and some test cases are duplicated and confusing.

Objectives

Migrate all the test cases to the implementation using ginkgo in e2enew and unify the code writing in e2enew. The ultimate goal is to improve functional test coverage and fix various

We tried to build dashboard docker image for ARM in #197 but failed. Since I could build dashboard docker image on m1 so there's nothing to do with the building process itself.

And since it stuck at yarn build, a large chance is free Github Actions machine is lack of memory since we also deploy QEMU to simulate arm environment, as [ref](https://stackoverflow.com/questions/59868448/rpc-error