ems-esp is rebooting twice a day

[tp1de]

@proddy
Hi Paul, I recognized that my ESP32 gateway is rebooting approx. after 11-12 hours. I can't see any reason why. Free memory is not the problem. Could it be that having http get requests per device every 15 sec per api is initiating the rebooting after a while?

It's not a major issue for me. I just miss some data for approx 90 secs because of my 60 tx-delay on startup.
Could I help to resolve this problem?

[proddy]

This is going to hard to find. It's probably the web code (API) but could you try and narrow it down to get it easily reproducible. For starters

• does it happen on WiFi or Ethernet or both?
• are you sure its rebooting and not just reconnecting? (i.e. uptime sets back to 0)?
• did you see similar issues with earlier v3 releases?
• does the same thing happen if you disable the other services like MQTT, NTP, SysLog?
• are you able to see from the logs when and where it crashes? is it always the same spot?
• if you increase the poll rate from 15 seconds to less, does it crash quicker?
• if you limit the number of API calls does it help?
• would it be helpful if there was an API call to group calls to prevent too many TCP connections?

[proddy]

well spotted @MichaelDvP !

[tp1de]

@proddy @MichaelDvP
I will test and report.

[tp1de]

@proddy @MichaelDvP

I installed the updated version with the fixed memory leak. Now it is getting worse.
I get a reboot now approx. every 13 minutes. It is a reboot of the esp32 and not just a connection lost for ems-bus.

I will try to watch the log if I can see anything ....

[proddy]

The fix @MichaelDvP did is valid. The async response buffer is only free'd after a send (https://github.com/me-no-dev/ESPAsyncWebServer#request-life-cycle).

[tp1de]

Now it is running since > 30 minutes. But I recognized the following in respect to web-ui system log:

• When loglevel is set to error I get 1-2 crc errors shown per minute
• When loglevel is set to debug then no crc errors are shown - that's strange -or ?

I switched off the web-ui system log and will wait if this has an influence.

[proddy]

When loglevel is set to error I get 1-2 crc errors shown per minute

That is really bad. What does show ems tell you from the console?

[tp1de]

What do you mean with bad? Too many crc errors?
But why aren't the errors shown in debug mode?
Actually the esp is running since 1:20 hours and system log-level is off:

[proddy]

What do you mean with bad? Too many crc errors?
But why aren't the errors shown in debug mode?

1-2 CRC errors every minute is not good. Are you sure you are seeing this?
Errors are also shown when the log level is DEBUG. If you're not seeing any probably there aren't any! It's hard to tell without real logs, screenshots etc

[MichaelDvP]

Reboots and crc-errors can be induced by a weak/unstable powersupply, thin cables, etc. Last in #72, and many times before.

[tp1de]

Reboots and crc-errors can be induced by a weak/unstable powersupply, thin cables, etc. Last in #72, and many times before.

Yes I am certainly aware, but I haven't changed anything on wirering and/or power supply.
I haven't had this "instability" in older versions for some months.
I will test in former builds if I will have the same behaviour.
0.9% crc errors on rx should not be a major problem as far as I can judge.

[tp1de]

The gateway rebooted again after approx 3,5 hours. I had log level on "off". After a reboot it is reset to "info".

[proddy]

I would ignore the CRC errors and how they are reported. It's normal to see CRC errors. Logs are only shown when the browser window is open and it's always set to INFO. So all normal there.

To find out why it's rebooting, try playing with the API calls (lower the number of calls, turn it off) until you narrow down what is causing it