Hi
I started playing around with a local cloud spark-server, now that’s it’s been released. I have a core connecting to a local server and it seemed to be working fine. Yay \o/
However, I quickly noticed a problem. Perhaps the easiest way to explain it is some steps for others to replicate the problem …
1: Start spark-server (node main.js).
2: Power up a pre-configured and ready to run local 'core.
The 'core connects, starts breathing cyan and otherwise works as expected. So far, so good.
3: Cause a disconnection between the core and server by any of the following means …
a) reset the core
b) power cycle the 'core
c) terminate the spark-server and re-launch it (optionally reset/power-cycle the 'core.)
At this point, the 'core can no longer establish a working connection (red flashing LED, then reboot, in endless loop) in spite of spark-server continuously reporting that it has …
server started { host: 'localhost', port: 5683 }
Connection from: 192.168.1.120, connId: 1
on ready { coreID: '53ff6d065067544853360587',
ip: '192.168.1.120',
product_id: 65535,
firmware_version: 65535,
cache_key: undefined }
Core online!
Connection from: 192.168.1.120, connId: 2
on ready { coreID: '53ff6d065067544853360587',
ip: '192.168.1.120',
product_id: 65535,
firmware_version: 65535,
cache_key: undefined }
Core online!
Connection from: 192.168.1.120, connId: 3
on ready { coreID: '53ff6d065067544853360587',
ip: '192.168.1.120',
product_id: 65535,
firmware_version: 65535,
cache_key: undefined }
Core online!
… and so on.
4: Cycle power on the core
No change
5: Power off core, terminate server, restart server, power up core
Still no change!
6: Wait for at least 3 minutes (maybe more) and repeat step 5. (Sometimes just waiting a long time while the core keeps rebooting will eventually bring it back online.)
Finally, the core connects again. But things repeat just the same, if we go back to step 3.
So in summary, the server reports that the 'core connected OK, but the 'core itself just barfs blood red laser light. Ouch. 
The reboot cycle LED flashing goes like so …
Solid cyan, fades, then rapid green flash, then rapid cyan for about half a second, then red flashing, which usually (but not always) appears to be, “SOS” in continental Morse code.
My guess is that a re-connection occurring before some time-out period has passed, results in a response being sent to the core that it is not understanding – or perhaps the timing is too fast on a local system or something? But I have no idea how to debug this problem further.
Any clues anyone?
If I can find a way to reliably reproduce this bug here, I’ll be much closer to fixing it.