Server response issues

Troubleshooting
1

Greetings from an enthusiastic new user.

I have an issue with unreliable responses from the server. As a baseline I can claim my core and connect to it via wifi. I can build code via the web IDE and verify and/or flash it to the core … at least I can sometimes.

When it all works it works as it should. More often than not I cannot even verify code via the web IDE.

I can always save code and load saved files, libraries etc. via the web IDE. When the issue presents, the following happens:

  • I press the verify button, and get the message
  • ‘saving firmware …’. Then I get the message
  • ‘Compiling code …’ but the verify
    button continues for a minute or more with the round arrows rotating until the IDE produces the message
  • ‘The server has failed to process the request on time. Please try again.’

Most of the time this is the result I get (9/10?). It makes it almost impossible to test code iteratively.

The core is mostly connected via wifi but occasionally will rapidly flash cyan while it renegotiates a connection. The core often fails to respond to ‘get’ requests as well even though it appears to be connected (breathing cyan). The get requests also typically timeout. This is consistent via my home network or 4G.

My internet connection and local wifi seem to be working fine.

Any thoughts on how I can test/resolve what is going on?

thanks, Arnie

2

Hey Arnie,

Thanks for posting! It would help troubleshoot if you could share some more details:

I know I have personally not had this much trouble with either the core talking to the cloud or with the webIDE so I imagine there’s something happening between you and the cloud. Let’s see if we can’t figure out what it is.

3

I’m running Windows 8.1, Chrome, Melbourne Australia (115.64.115.201).

I have a log of errors from a Google spreadsheet that was polling temperatures every 10 minutes. It runs for about four days and there are blocks of successful ‘get’ requests and then large blocks of failed requests. I can share that if it is helpful.

Late afternoon/evening and again this morning (Sunday and Monday in Australia) it has been working reliably.

thanks, Arnie

4

An update.

Cloud requests/communication have been reliable all week until now. I cannot verify/flash my core, and get requests mostly time out. I don’t seem to have any issues with wifi or my broadband connection.

5

I would make sure that your access_token has not expired.

http://docs.spark.io/api/#authentication-list-all-your-tokens

6

The access token is ok. Things seem to be working again but for four hours this morning I had trouble. I haven’t changed anything. It’s great when it works, very frustrating when server responses time out for large slabs of time.

7

@arnlew, sadly I’m out of ideas for now. Do you have the resources to run a long-running latency-checking script on an always-on or nearly always-on computer that is connected to the same network as the core? Perhaps there’s something going on upstream that is intermittent.

If the answer is yes I can probably write something in a few days that will spit out a log file for us to analyze.

8

I’m not sure I can answer your question because I don’t know what ‘long-running latency-checking’ means as such. I am familiar with arduino IDE and coding, I’m working on learning javascript, but I haven’t installed the CLI. I’m slowly getting my head around GET and POST requests but a lot of this is new stuff for me.

My machine is a tablet/PC device so is often on the road with me for work. I plug it into a docking station at home.

I can run tests while things work, and then again when I have server response issues if that helps.

Thanks for your time.