Spark Core no longer communicating with the cloud

I was testing the reliability of my code and left my Core plugged in and running last night. When I woke up the code was still running correctly, but I could no longer communicate with it. I reset it, same result. I then did a factory reset, set the WiFi credentials via Serial, still can’t connect. In all instance it is obtaining an IP from DHCP AND the LED is breathing cyan as if it was connected. I cannot flash or interact in anyway with the Core via the cloud. I was hoping there were cloud issues, but the status page shows no issues. What could be the problem!?

EDIT: I also want to add that I cannot ping the Core from another computer on the local network. I don’t know if I ever could before, but I cannot now. I know it gets an IP because every time I delete the lease and reset the Core it grabs an IP again.

Hey Mnetwork, do you have the local build tools setup? Ie, can you write and flash code locally (over USB) or have you mainly been using the WebIDE or SparkCLI to send code?

I don’t usually use them, but I don’t see why those wouldn’t work. The Core executes code and communicates serially just fine. The issue is purely with cloud communication.

Oh. I was just concerned about you being able to program it going forward.

What happens if you flash the example HTTPClient code on there and then view the serial output? Does it appear to connect to the time server? (see https://www.spark.io/libs/53dbe5bfd9132eb007000b3d/tab/application.cpp)

I will have to try this tomorrow. I just received an outage resolution email. Could this be related?

It definitely could be. Let me know if you are still having this issue when you get a chance

Do we know what the actual problem was? There wasn’t much detail. I won’t have access to my Core until tomorrow evening.

Hi @mnetwork,

Sorry about that! We had a brief database outage on one of our boxes overnight on Sunday, and although things mostly healed as expected, a few devices were left in a temporary ghosted state which made them inaccessible through the API. I’m working on fixing the self-healing behavior so we can more fully automatically recover in those scenarios. In the meantime I restarted the weakened service, so all should be back to normal, and we’ll keep a close eye on it. I hope that helps!

Thanks,
David

@dave Well that sounds exactly like my issue! I’ll test my Core when I get home from work and report back. Thank you!

@Dave All good now!

1 Like