Ben Fradet
Ben Fradet
It may happen that both Kafka's and Hbase's zookeeper will try to bind on the same port which obviously will fail either suite. AFAICT, right now we can only force...
I think it'd be better to create the super API key directly and not have someone create it manually since the DDL is already executed when you launch the server...
Similarly to the schema service, it would be nice to be able to, for example, retrieve multiple keys from one request: ``` /api/apikeys/keys/an-uuid,another-uuid ``` Same goes for vendor prefixes.
Relatedly to #60, I think it would be more RESTful to move away from requests like: ``` /api/apikeys/keygen?vendor_prefix=com.acme ``` To something like: ``` /api/apikeys/com.acme ``` Removing the `keygen` keyword since...
As a spin off of #60, it'd be nice to change the API key generation service endpoint from: ``` /api/auth/keygen ``` to: ``` /api/apikeys/keygen ``` In order to be more...
It'd be cool to set up an endpoint where you could get the owner given an uuid or uuids given an owner like so: ``` /api/auth/apikeys/keys/some-uuid ``` You would get...
Spin off of #42 for requests like: `/api/schemas/vendor/name/format/1-0` which should return: ``` json [ { //schema with version 1-0-0 }, { //schema with version 1-0-1 }, //etc ] ```
The situation has been getting better wrt Spark jobs running on spot instances in EMR recently (https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-spark-configure.html) so it might be interesting to support them.