86.4 CRASH / Will NOT boot

Attempted to upgrade (as I have done several times prior) via UPGRADE (Hassio menu). Crashed, failed last night. This morning, for my Rasp Pi 3 B+, FORMATTED the SD card (full format, not ‘Quick’), wrote new hassos_rpi3-1.13 image to SD card/Validate (just as I have done before) installed in Pi, powered up. Waited about 1 HOUR, would not come up in my browser. Turned Pi power OFF, waited, reapplied power, still nothing after 30 min. Downloaded a fresh IMAGE (hassos_rpi3-1.13.img), AGAIN FORMATTED the SD card (full format), wrote the newly downloaded hassos_rpi3-1.13 image to SD card/Validate (just as I have done before) installed in Pi, powered up. Waited about 1 HOUR, would not come up in my browser.

UPDATE: Feb-2, 2019… Yesterday, PM, on a hunch, I decided to take a look at my modem logs, good streams then JUNK,GARBLED, then OK again. (Did I mention that I hate my ISP?) On the end of the line here and AT&T will not guarantee ANYTHING! My guess is that while in the process of updating, JUNK was entered into the stream, causing the crash. Finally, I unplugged the modem for 5 min, powered up, did a FRESH download of the latest image, re-flashed the SD card, loaded the Rasp Pi…same configuration.yaml, same lovelace config, reloaded everything the same. EVERYTHING came up exactly as it should!
Thanks to everyone for trying to help me…I appreciate ALL your comments.

Hi, your issue may be caused due to an error in your configuration file. In case you made changes your configuration.yaml should always check if your configuration.yaml file is still valid before rebooting HA. To do so, go to the menu configuration / General / Check config in the UI. Even a simple typo can prevent HA from booting.

Furthermore, new HA releases can contain breaking changes. In such a case you need to fix your configuration.yaml otherwise HA won’t boot. Those breaking changes are published in the change log, see:

Were do you store your configuration.yaml? I assume on an a mounted usb device? You could try to fix it and reboot HA. An other way is to temporary remove parts (put a # in front of the lines) from the configuration and then see if it will boot again. Later on you can add those parts back and check if the configuration is valid using the UI.

You probably don’t have ssh access to your HA. It could save you a lot of trouble in case HA does not boot. Using SSH access your can check the logs and fix configuration changes.

If you do have ssh access, you can check your configuration using:

$ hass --script check_config

See:

If you need more help, please share your configuration.yaml file.

My advise for next time: don’t start with a new image. My experience is that the fix is in most cases very simple and it does not have to take so much time, but you need to have access to the HA logging (SSH access needed) to see what went wrong. See:

I hope this helps you!

Not true on a new install.

1 Like

I had the same problem, found three places in customization.yaml where there was double underscore in the entity names. Changed this to single underscore. After that upgrade worked like a charm!

Do you still have some double underscores left?

1 Like

Thanks for trying to help (see my edit above)

Thanks for trying to help (see my edit above) (and yes, a few upgrades ago, same underscore issue)

Thanks for trying to help (see my edit above) , last re-load was to upgrade and remove the last of the MQTT broker entries.