Dashboard control error + Rules empty

1

This evening our Sharptools dashboard stopped responding, giving an error message on some taps, but mostly ignoring any input. After looking around, it seems as if all rules have been deleted. The rules are still in the list, but the contents are non existent. Yet, the dashboard still receives updates from Homey through webhooks as they are initiated from Homey. The rules processing these webhooks are still active, yet it has no content when I try to open the rule.

Is this a problem on Sharptools side, or have I done something “intelligent”?

Please help ASAP.

2

Yikes! I dont know that I can help, so I’m mainly replying in case it gets you more visibility.

I’d log out / in SharpTools, try another browswer, clear cookies.

3

I’ve seen 2-3 reports of something similar with different resolutions in each case.

  • @Fred_Aydlett experienced this and rebooting his ATT router fixed things
  • Temporarily remove or disable any blocking tools (or whitelist SharpTools)
    • Ad blockers, VPNs, Pi Hole, and other Network Security tools
  • Others resolved themselves in < 24 hours, but it’s unclear why
    • I suspect some sort of common block / filter list was incorrectly updated which Ad Block tools or ISPs use and was later corrected

In each case I investigated, the initial real-time connection was established which meant dashboards were still visible, device statuses were updating, and rules continued to run in the background.

And some API calls like healthchecks were going through as expected, but almost all other API requests were being blocked. This means that sending commands from a dashboard, viewing rule details, and even loading recent invoices would fail.

–

By blocked, I mean that we would never even see the API requests hit our platform. It was like the requests were either being blocked by the browser or at the network level.

One community member even exported a recording of the network traffic from their browser and we could see that the realtime connection and healthcheck API calls went through, but the device command and other API calls were completely missing. In that case, there was suspiciously a VPN Extension in Chrome intercepting all the traffic, but the issue resolved itself in a few hours before we could troubleshoot further.