Most of the tests that currently cover core functionalities depend heavily on (and sometimes test) internal implementation characteristics, instead of testing whether given implementation really produces desired outcome (treating its implementation more as a black box).

This is very problematic for eventual internal improvements which occasionally we want to introduce, to make Framework more s

Move all Dockerfile samples to non-root user

Expected Behaviour

As an OpenFaaS end-user, I want to run all samples on my OpenFaaS cluster, even with the new non-root feature enabled.

Current Behaviour

Add a performance test that measures the time it takes Dapr to save data into a state store.

The important piece that is being tested is the processing that Dapr does before saving to the state store. This test should not benchmark the component itself.

For that purpose, a baseline test scenario would have an app that saves a 1kb payload to Redis, followed by a test app that uses Dapr to sav

Description:

I'd like to be able to specify the IAM role Path on either a function or globals.

Observed result:

Able to specify Path on a Function or Globals

Expected result:

Path is set on the implicitly created IAM roles

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

Missing functionality to update min/max cpu/mem of an exiting ENV. Currently, to update, we are forced to delete and re-create the env. This also has a risk of dependent functions being erratic. It also de-couple resource utilization independent of function creation and be useful to update the resource in a quick to respond situ

The typescript runtime was recently added to openwhisk but we do not yet have docs for the runtime and it is missing from runtimes.json.

We need a doc like https://github.com/apache/incubator-openwhisk/blob/master/docs/actions-nodejs.md for typescript functions.

Adding the runtime to the runtime manifest can be done per https://github.com/apache/openwhisk/blob/master/docs/actions-new.md#the-

this will very quickly get out of hand if we allow it, i realize we don't have principles written down anywhere but to date one of them has been that binaries 'just work' without having to do any configuration. currently a user must set FN_MAX_REQUEST_SIZE, FN_MAX_RESPONSE_SIZE -- neither of these should be required to run.

test case:

cd tests/fn-system-tests/
go test -v

this

https://emotion.sh/

After this change in Material UI: mui-org/material-ui#6115

We currently only test our autoscaling behavior for concurrency and rps. We should add an e2e test that also asserts we properly scale on `