the mock backend needs to have a response, that google's FCM lib (which is used in UPS), needs to understand. Otherwise, well ... sure, no luck.

In case you are using wiremock, please do something like:
curl -X POST --data '{ "request": { "url": "/fcm/send", "method": "POST" }, "response": { "status": 200, "fixedDelayMilliseconds": 2000, "jsonBody": {"multicast_id": 108,"success": 1,"failure": 0,"canonical_ids": 0,"results": [{ "message_id": "1:08"}]}, "headers": {"Content-Type": "application/json"}}}' http://localhost:9090/__admin/mappings/new
so that the loclahost:3000/fcm/send thing sends a meaningful response.

-Matthias



On Wed, Jan 25, 2017 at 5:38 PM, Jose Miguel Gallas Olmedo <jgallaso@redhat.com> wrote:
No, it just sends "null". But that does not seem to be an issue for the first app :/

On 25 January 2017 at 15:56, Matthias Wessendorf <mwessend@redhat.com> wrote:
and the tool that intercepts the requests also sends back some meaning full reponse ? 

like described here:



On Wed, Jan 25, 2017 at 3:49 PM, Jose Miguel Gallas Olmedo <jgallaso@redhat.com> wrote:
Yes, the proxy is up and the -Dcustom.aerogear.fcm.push.host is set, the notifications are actually being sent.

On 25 January 2017 at 15:44, Matthias Wessendorf <mwessend@redhat.com> wrote:
endpoints for thje /fcm/send mock are setup ? 



On Wed, Jan 25, 2017 at 2:07 PM, Jose Miguel Gallas Olmedo <jgallaso@redhat.com> wrote:

---------- Forwarded message ----------
From: Jose Miguel Gallas Olmedo <jgallaso@redhat.com>
Date: 25 January 2017 at 13:00
Subject: Mysterious warning when sending push
To: AeroGear Developer Mailing List <aerogear-dev@lists.jboss.org>


Hi all,

I have some phenomenon here. I have created 4 variants with 100 fake tokens and sending 1 push through the mock FCM server. What I got is 400 (401 to be exact) receivers so it seems to work OK.

These are the 4 variants (the order is important):


However ONLY on 2nd to 4nd variant I got this warning:

​

The first variant in the list is not getting the error. I created them in order (0 - 3) so this does not matter because the 0 (second in the list) is receiving the warning but the 1 (first in the list) is not.

​Look what happens if I delete the first variant in the list and send another notification:

​
​(Now the first in the list is variant 0, the former 2nd in the list)



​This is no longer receiving the warning!​
​

​
​
So what do you think? Bug? It does not seem to be related with the fact that I am using the FCM proxy... or does it? Maybe I'm missing something obvious.

Cheers,

​
--
Jose Miguel Gallas Olmedo
Associate QE at Mobile Team, Red Hat



--
Jose Miguel Gallas Olmedo
Associate QE at Mobile Team, Red Hat



--
Project lead AeroGear.org



--
Jose Miguel Gallas Olmedo
Associate QE at Mobile Team, Red Hat



--
Project lead AeroGear.org



--
Jose Miguel Gallas Olmedo
Associate QE at Mobile Team, Red Hat



--
Project lead AeroGear.org