I meant that Device Activity is not working
If I manually feed a message and then re-authenticate with Dave's flow, the activity on the device works for a few minutes
1 Like
As an experiment, disable your trigger node and redeploy
Then stop/restart Node-RED
See if it works for more than 2 or 3 messages BUT DO NOT redeploy while you are testing
I currently believe that the fault only appears if it is initialised more than once after a Node-RED restart.
1 Like
Just tried that - it only performs or responds to 3 Alexa commands before she stops reacting (hiss, boo).
OK
How about disable your existing Init node
Add in another one - do not connect it to anything - make sure it's set to auto init
Deploy
stop/restart Node-RED - DO NOT do any deploys while testing
What happens?
1 Like
At my end - I ran a test sequence of my automated Simon says 1 to 69 over 20 minute period
All worked prefectly - after message 53 - i re-initilised - it stopped working at message 58 and only restarted at message 63
At my end, It definitely seems to work as long as I never re-initialise
1 Like
Hi Simon - I agree with your findings.
It seems to be much more reliable when doing that - tried it over 10 times, all worked.
Note: My Alexa node is set to refresh every 3-days (which I think is the default).
What does '0' mean - never do a refresh??
Just changed my setting to '0' to see what happens.
AFAIK 0 means never refresh.
Don't get me wrong - its still broken - but I think we might be zeroing in on what causes it to go into failure mode 
before all this happend - I was like you - used to send an init in every few hours
And I think this explains why some people are having the issue and some people not
So I'm now running my Alexa node in a separate instance that only has one flow (which I'm locking) and I use MQTT to send received events off to my other NR instances that need the data.
Hopefully, this will act as a working workaround until a proper solution is found by the wizards
Of course, I think I've found a solution about 10 times in the past few weeks so I'm probably wrong about this one as well
But it currently matches up with cause and effect
2 Likes
Well it certainly seems more reliable if I don't 'init' it again.
Certainly need to stop/start Node-RED.
Performed over 24 successful commands at the time of writing this post.
EDIT: Over 1hr and 60+ commands and the 'OnDevAct' is still working.
EDIT: Just noticed an update for the node has been posted.
EDIT: 2 hrs and still working fine.
3 Likes
Watching last night's strickly at the moment but will test later
-
removed Alexa event node and links
-
recreated another Alexa event node,connected to each switch node like the previous one,
-
disabled refresh with 0 in alexa account,
-
restarted NODE-red
and everything has been working correctly for about an hour.
1 Like
See my comment on GitHub
1 Like
Disabled refresh and installed latest version - so far so good 
2 Likes
Just checked this morning - 'OnDevAct' is still working. That's over 13 hrs of use.
So it would seem that performing an 'init' is in some way related to the instability problem.
Edit: Just checked again - still working - that's nearly 18 hrs.
Edit: Nearly a whole day (24 hrs) of use in my NR flow - still working.
2 Likes
I also for about 9 hours ,everything works smoothly.
I have the same problems again 
Same here - my 'OnDevAct' has stopped working this afternoon !!!!
1 Like
Yep - confirmed -stopped in past few hours
Reported it on Github
2 Likes
Just to jump on the bandwagon - also not working here 
1 Like
