Clock#Time not logging frequently enough

[andy]

Sometimes my pump relay does not trigger on the Clock#Time event. It seems there are times that more than a minute elapses before ESP reports a new time. For instance, I will see 13:59 registered in the log, but it will miss registering 14:00. It will register 14:01, but by then I've missed the 14:00 event I needed to trigger a relay.

I'm currently getting around it by setting two rules to turn on and two to turn off, each of the pairs separated by 1 minute. That way, if the first rule doesn't hit, the second one should. But it takes up more memory and is a bit of a kludge - I'd like to figure out how to configure my way around it.

Any ideas?

[Martinus]

The main program loop checks system time every second, unless a sub process takes more time. There must be some reason that a sub process takes more than a minute doing some other stuff, stalling the main loop for too long.

Does this unit have many tasks?
What's your message delay setting?
System load?

[andy]

Does this unit have many tasks?

I'm measuring a DHT11 and a ds18b20. I've got an I2C bus with an ADS1115 and six relays on it.

TASKS:

• Uptime
  Wifi Strength
  DHT11 Temp & Humidity
  ds18b20 temperature
  0.5 - 4.5 VDC signal on ADS1115

RULES:

• On boot set pcfgpio 1 & 2 to 1
  On Clock#Time = 8.00 set pcfgpio 1 & 2 to 0
  On Clock#Time = 8.01 set pcfgpio 1 & 2 to 0
  On Clock#Time = 12.00 set pcfgpio 1 & 2 to 1
  On Clock#Time = 12.01 set pcfgpio 1 & 2 to 1

What's your message delay setting?

Message delay is 15000 (required by ThingSpeak)

System load?

Sitting at 100%, even with serial and web logging set to 0.

[Martinus]

Check your task delay settings. It may help to set them to larger values. A starting point could be 30 x "number of tasks reported".
So if you have 4 tasks sending data to ThingSpeak, set all delays to a minimum of 4 x 30 = 120 seconds.
System load should drop from 100% and rules may run more reliable.