MB data logger stuck - attempt to write a readonly database (8) - since 14074 **solved**

[Gyvate]

Hi Boris,
since update 14074 my MB is stuck.
It keeps on connecting, disconnecting but is stuck with this database error.
Reboot doesn't change anything.
logger (18.11.2020 14:18:40): DB cannot prepare select on 'wind0avgwind_year' (after 100 tries in 9217ms): attempt to write a readonly database (8)
logger (18.11.2020 14:18:50): DB cannot prepare select on 'wind0avgwind_year' (after 100 tries in 9213ms): attempt to write a readonly database (8)
logger (18.11.2020 14:18:57): child process received timeout signal (14).
logger (18.11.2020 14:18:57): disconnect station 0 (GW1000): exit
logger (18.11.2020 14:18:57): disconnect station 1 (GW1000): exit
logger (18.11.2020 14:18:57): disconnect station 9 (System Data): exit
logger (18.11.2020 14:18:58): data logger stopped.
logger (18.11.2020 14:18:58): databases disconnected.
logger (18.11.2020 14:18:58): parent process to be terminated.
logger (18.11.2020 14:18:58): databases connected. Update interval 3600 seconds.
logger (18.11.2020 14:18:58): data logger (version 5.1d, build 14074) started.
logger (18.11.2020 14:18:59): connect station 0 (GW1000 via TCP/IP).
logger (18.11.2020 14:18:59): connect station 1 (GW1000 via TCP/IP).
logger (18.11.2020 14:18:59): connect station 9 (System Data via Plug-in).
logger (18.11.2020 14:19:01): waiting for sensors (wind0, rain0, th0, thb0) to provide data.
logger (18.11.2020 14:19:01): services paused.
logger (18.11.2020 14:19:09): sensors are now providing data.
logger (18.11.2020 14:19:09): services started.
logger (18.11.2020 14:19:18): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9213ms): attempt to write a readonly database (8)
logger (18.11.2020 14:19:27): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9222ms): attempt to write a readonly database (8)
logger (18.11.2020 14:19:36): DB cannot prepare select on 'th0temp_day' (after 100 tries in 9200ms): attempt to write a readonly database (8)
logger (18.11.2020 14:19:45): DB cannot prepare select on 'th0temp_month' (after 100 tries in 9197ms): attempt to write a readonly database (8)
logger (18.11.2020 14:19:55): DB cannot prepare select on 'th0temp_month' (after 100 tries in 9280ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:04): DB cannot prepare select on 'th0temp_month' (after 100 tries in 9267ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:13): DB cannot prepare select on 'th0temp_year' (after 100 tries in 9202ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:23): DB cannot prepare select on 'th0temp_year' (after 100 tries in 9271ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:32): DB cannot prepare select on 'th0temp_year' (after 100 tries in 9206ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:41): DB cannot prepare select on 'th0hum_day' (after 100 tries in 9234ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:50): DB cannot prepare select on 'th0hum_day' (after 100 tries in 9263ms): attempt to write a readonly database (8)
logger (18.11.2020 14:20:59): DB cannot prepare select on 'th0hum_day' (after 100 tries in 9194ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:09): DB cannot prepare select on 'th0hum_month' (after 100 tries in 9224ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:18): DB cannot prepare select on 'th0hum_month' (after 100 tries in 9215ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:27): DB cannot prepare select on 'th0hum_month' (after 100 tries in 9232ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:36): DB cannot prepare select on 'th0hum_year' (after 100 tries in 9201ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:46): DB cannot prepare select on 'th0hum_year' (after 100 tries in 9199ms): attempt to write a readonly database (8)
logger (18.11.2020 14:21:55): DB cannot prepare select on 'th0hum_year' (after 100 tries in 9212ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:04): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9205ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:13): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9265ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:22): DB cannot prepare select on 'thb0seapress_day' (after 100 tries in 9195ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:32): DB cannot prepare select on 'thb0seapress_month' (after 100 tries in 9243ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:41): DB cannot prepare select on 'thb0seapress_month' (after 100 tries in 9262ms): attempt to write a readonly database (8)
logger (18.11.2020 14:22:50): DB cannot prepare select on 'thb0seapress_month' (after 100 tries in 9242ms): attempt to write a readonly database (8)
logger (18.11.2020 14:23:00): DB cannot prepare select on 'thb0seapress_year' (after 100 tries in 9630ms): attempt to write a readonly database (8)

[Gyvate]

"Progress" report:
I did a cold start (power cycle) - no change
tried to restore several backups - no change
tried to restore the database from my test/backup machine which is an identical clone of my prod system - no change
(the backup machine still runs on 14072 and the newest version on reboot is now deactivated).
So either the hardware is broken (unlikely) - or the 14074 introduced a bug which keeps my system down.

[Gyvate]

Got it working again with "Rebuild on reboot".

Looks like the RAM DB was corrupted and messing up everything.
I had similar cases before, but then the system would just purge the RAM DB and from scratch (RAM DB).
Somehow this step didn't happen anymore.
Now I have to manually enter the missing data (6 hours) from the "mirrored" database. 12 entries per sensor and station.
Could have been worse.

But it's still not fully clear to me why a simple restore wouldn't work.
The only explanation left to me that in such a case I have to connect to the MB via SSH,
delete the DB and the RAM DB and then only start the restore.