Anybody else experiencing erratic overlay behaviour?

[Lorenzi70]

I have my allsky (PI4 , raspbian 32bit, ZWO 224MC) that sometime stops adding the overlay on the images for a day or two, than get back working again. It happened already few times. Rebooting and/or removing (resetting)/adding back the modules does not work.
Only modules selected in my configuration are BME280 (see my former post), Cloud(MXL90614) & Overlay data (in the order Top>Bottom).
Anybody else having issues? Could it be there are troubles with the variable database generation/reading and somehow this is preventing the overley script to work?

Thanks
Marco

[EricClaeys]

@Alex-developer, any ideas?

[Alex-developer]

I would need to see the allsky.log file to see whats going on.

The module must be enabled as if it were disabled it could only be re enabled manually.

Can you attach the allsky.log file and I will see if I can see whats going on

Thanks

Alex

[Lorenzi70]

Thanks for jumping in Alex. After a while I found out the location of the log (i.e. /var/log/allsky.log) but the size is almost 60MB, too large to attach here. but I put it into my dropbox"https://www.dropbox.com/sh/m5qu6s38pmh7fge/AAA7P-iv11iYsWNeyDVKRSDka?dl=0")
I have checked the log for error messages, in particular related to overlay.py. Got few of them:
1) lost comm with the I2C sensors
May 26 22:17:37 allskybali allsky[4360]: ERROR: No I2C device at address: 0x5a
May 26 22:17:37 allskybali allsky[4360]: ERROR: Unable to download de421.bsp. cannot download https://ssd.jpl.nasa.gov/ftp/eph/planets/bsp/de421.bsp because <urlopen error [Errno 113] No route to host>

May 26 22:18:12 allskybali allsky[4311]: ERROR: Module allsky_overlay.py will be disabled, it took 77.09s max allowed is 30s

I understand that the overlay module will crash if errors in communication by any added device or issues of too many errors.
Funny also got a lot of non critical errors related to the colors:
May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${DATE} ${TIME}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Exp: ${sEXPOSURE}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Gain: ${GAIN}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Sensor: ${TEMPERATURE_C} C' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${CAMERA_TYPE} ${CAMERA_MODEL}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${MOON_SYMBOL}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${DAY_OR_NIGHT}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'illumination:${MOON_ILLUMINATION}%' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Sun rise: ${SUN_SUNRISE}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Sun set: ${SUN_SUNSET}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Dew ${BME_DEWPOINT}C' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Baro ${BME_BAROPRES} mB' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Amb ${BME_TAMBIENT}C' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Hum ${BME_HUMIDITY}%' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Sky ${CLOUDSKY}C' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${CLOUDCOVER}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'AstroBali Weather' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field '${CLOUDMODEL}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Moon rise: ${MOONRISE}' Is NOT valid - Defaulting to white May 24 10:44:20 allskybali allsky[26001]: ERROR: The colour '' for field 'Moon set: ${MOONSET}' Is NOT valid - Defaulting to white

In the overlay menu I found nothing strange with all colors porperly allocated by default, I have however manually set white color on both font and stroke (desabled) in the default menu and the error messages now disapperead..

Regards
Marco

[Alex-developer]

Thanks for jumping in Alex. After a while I found out the location of the log (i.e. /var/log/allsky.log) but the size is almost 60MB, too large to attach here. but I put it into my dropbox"https://www.dropbox.com/sh/m5qu6s38pmh7fge/AAA7P-iv11iYsWNeyDVKRSDka?dl=0") I have checked the log for error messages, in particular related to overlay.py. Got few of them: 1) lost comm with the I2C sensors May 26 22:17:37 allskybali allsky[4360]: ERROR: No I2C device at address: 0x5a May 26 22:17:37 allskybali allsky[4360]: ERROR: Unable to download de421.bsp. cannot download https://ssd.jpl.nasa.gov/ftp/eph/planets/bsp/de421.bsp because <urlopen error [Errno 113] No route to host> May 26 22:18:12 allskybali allsky[4311]: ERROR: Module allsky_overlay.py will be disabled, it took 77.09s max allowed is 30s
....

Marco

Few things here

• The colour error has been fixed in the next release and can be ignored.
• The de421.bsp error can be ignored unless you are using any planet data. Currently this file is downloaded to /tmp but in the next release i have moved it into allsky so it will only ever be downloaded once.

The error thats concerning is the overlay taking too long as that will automatically disable it. When you get the problem can you log into the pi and do a top and see what the load average is please

Thanks

Alex

[Lorenzi70]

Thanks Alex. Ok I will keep an eye if I get the problem again, looking at the past videos it happens once during the transition night/day (hence with the RB probably loaded with othe activities) but also during the morning with presumably less cpu load.. since last three days no issues, I have also welded the external sensors now and this should fix the sensor connectivity issues I had..
Regards
Marco

[Alex-developer]

Thanks Alex. Ok I will keep an eye if I get the problem again, looking at the past videos it happens once during the transition night/day (hence with the RB probably loaded with othe activities) but also during the morning with presumably less cpu load.. since last three days no issues, I have also welded the external sensors now and this should fix the sensor connectivity issues I had.. Regards Marco

The other thing I noticed is that it appears you are saving daytime images? Make sure there is a long enough delay between the images to allow the overlay to run, this is not really an issue at night as the exposures are longer. If you don't allow enough time for the overlay to run then you can end up with multiple overlay processes running which will slow the pi down

[Lorenzi70]

Thanks for the feedback Alex, I believe I left enough time for the overhedas but it is a very good reminder..
I still have some issues with custom modules (I will open a post) but it seems that since I hardwrired the sensors the erratic behaviour is no more. I will close this topic as answered :)
Thanks
Marco