AlexxIT
Lots of Cloud 504 errors
I'm having lots of Cloud errors, the 504 on
I'm wondering if the time out period is to short before the error is issued.
I've done tracert to ewelink.cc (which is a Cloudfront 162.159.137.9 IP). I'm in the U.S.A. so I gather that is where my Sonoff s31 are sending data.
The other piece of information is that it actually tasks 42 seconds for that trace route to complete. (timed on a Stop Watch watching monitor screen)
Ping results are faster.
Everything works, data is gathered, automations work, just ton of these errors in the log everyday home-assistant_sonoff_2023-11-16T18-07-47.809Z.log
just wanted to report I am seeing the exact same behavior as well. Started yesterday and continues into today filling up logs.
I think this is a cloud issue which we as users have no control over. That is why I'm asking about a longer time period before throwing the error that the devices can't phone home. Just the most important reason I try to avoid devices that want you to use their cloud tied usage. I'd hate to be locked out of my home because the door locks or the garage doors can't make it's check in to home base and operate.
Trace route is not always - let's say correct - it involves additional information requests from devices "on the way there". simple ping is with info you require, sometimes trouble is somewhere in the middle (some devices on the way have a higher load). Also sometimes you can use ping with larger packet size. You could try testing using WinMTR (free for download and use) application - running for a little while it would offer some additional info.
So I started a experiment today. Now that HA has a ping/icmp integration, I've set it up to check the eWeLink sites used in the Add-on. It sends 1 ping packet (down from the 5 pings) every 5 minutes (default time to send). It uses the info I found in the custom_configuration for eWeLink. Hopefully it will provide some info, I did find out that the site used for the API will not accept Trace Route or Ping inquires, so that is interesting. The site used for WS does though so that is interesting. I'm using the US location ones since I'm in the U.S.A. Pings average around 70 to 75 ms. So for others this could be a great debugging tool, time will tell. More info: Yes and it gave me different IP addresses than what was listed before. Interesting!
The issue happens mostly overnight, between the hours of 1 AM to 6 AM. So it has to be maintenance of the Cloud (Servers, Network Switches, Routing changes, Software, etc). Otherwise runs perfect during the day. Just clear the issue with a restart of HA.