Hello Antoine,

I am running the test again, w/ the 1.0.0.Beta1 branch. It's faster now. I am getting now 100ms per request, I am in the middle of importing a 100k devices (48.000 devices are now updated). 

Also, I still want to write a special endpoint that takes a (large) JSON file, and imports all the device metadata in one step.

greetings,
Matthias


On Mon, May 26, 2014 at 10:59 AM, Matthias Wessendorf <matzew@apache.org> wrote:
Hello Antoine,

I noticed the same issue, I will be looking into this, so that we can have a better 'batch' import.

Thanks for reporting!
Matthias


On Wed, May 21, 2014 at 4:33 PM, Matthias Wessendorf <matzew@apache.org> wrote:
hey Antoine,

thanks for explaining - I will take a look at this:

I will think about making the endpoint completely async (there is no real benefit in returning a JSON of the device metadata anyways + do some overhaul on the actual database)


On Wed, May 21, 2014 at 3:33 PM, A577127 <antoine.matyja@worldline.com> wrote:
Uh wrong translation  I mean register tokens. We wrote a java sender that
registers lots of token (we wanted to register 1M tokens) with the REST api
http://aerogear.org/docs/specs/aerogear-push-rest/DeviceRegistration/

But it was very slow (~1/s), looks like the database slows the thing by
making big requests (huge joins ?).



--
View this message in context: http://aerogear-dev.1069024.n5.nabble.com/aerogear-dev-GCM-limit-with-1000-registrationIDs-tp7858p7887.html
Sent from the aerogear-dev mailing list archive at Nabble.com.
_______________________________________________
aerogear-dev mailing list
aerogear-dev@lists.jboss.org
https://lists.jboss.org/mailman/listinfo/aerogear-dev






--
Matthias Wessendorf

blog: http://matthiaswessendorf.wordpress.com/
sessions: http://www.slideshare.net/mwessendorf
twitter: http://twitter.com/mwessendorf



--
Matthias Wessendorf

blog: http://matthiaswessendorf.wordpress.com/
sessions: http://www.slideshare.net/mwessendorf
twitter: http://twitter.com/mwessendorf