These are chat archives for nightscout/beta

27th
Mar 2015
Jim Sifferle
@jimsiff
Mar 27 2015 20:01
@bewest @ktind @jasoncalabrese I can't find anything on how to setup MQTT. Based on the c-r-m code and my CloudMQTT setup, I'm guessing that I set the MQTT_MONITOR variable in c-r-m to the URL in the CloudMQTT instance details?
The URL CloudMQTT recommends is mqtt://user:pass@hostname:port. The Configure page uses tcp://, while @ktind asked to switch from tcp:// to ssl:// in nightscout/nightscout.github.io#49.
Kevin Lee
@ktind
Mar 27 2015 20:22
Screenshot_032715_032140_PM.jpg
does that help? =)
ignore the ?clientId=XXX option
that is me overriding the default behavior for testing
Jim Sifferle
@jimsiff
Mar 27 2015 20:23
Perfect... thank you!
Kevin Lee
@ktind
Mar 27 2015 20:24
Note that cloudmqtt has both an SSL and TCP port
the instance details overview only shows you the TCP port
to get the SSL port you have to go into the Mgmt interface
Jim Sifferle
@jimsiff
Mar 27 2015 20:25
found it.
Kevin Lee
@ktind
Mar 27 2015 20:26
This message was deleted
cool
Jim Sifferle
@jimsiff
Mar 27 2015 20:26
is connection limit defined as the # of endpoints talking to the message broker? So, the website is 1, the uploader is another...
Kevin Lee
@ktind
Mar 27 2015 20:27
Correct
and hopefully soon.. the native monitors will be more
Jim Sifferle
@jimsiff
Mar 27 2015 20:27
and if c-r-m uses MQTT for client updates at some point, each web client would be another...
Kevin Lee
@ktind
Mar 27 2015 20:27
Also, keep in mind that I don't think Raw values are in the format that the CRM expects for MQTT
Jim Sifferle
@jimsiff
Mar 27 2015 20:28
right... so at the moment, MQTT is sending separate documents for the ISIG data and the EGV data?
Kevin Lee
@ktind
Mar 27 2015 20:28
We could fill those 10 easily =)
Yeah
It is packaging everything up into a single download object
Jim Sifferle
@jimsiff
Mar 27 2015 20:29
:+1: I will probably setup a second db for testing MQTT
Kevin Lee
@ktind
Mar 27 2015 20:29
I usually just use another collection
Jim Sifferle
@jimsiff
Mar 27 2015 20:30
yeah I should just do that. I already do for our primary data. Allows us to use a common treatments collection for a variety of rig configurations