After updating to the latest core HA, the system would not start, except in safe mode.
Managed to get the system running with a backup of one week ago.
All seems to work, except that the recorder does not work, with a message in the logs:
''' The system could not validate that the sqlite3 database at //config/home-assistant_v2.db was shutdown cleanly '''
At present no access to the history, no access to energy data, no access to any historical records.
I did not find a .CORRUPT file, just the normal database stuff.
Or do you mean it IS CORRUPT (rather than being named ".CORRUPT")?
/config ls -l home-assistant*.*
-rw-r--r-- 1 root root 5124157440 Jun 23 08:55 home-assistant_v2.db
-rw-r--r-- 1 root root 98304 Jun 23 08:58 home-assistant_v2.db-shm
-rw-r--r-- 1 root root 4194192 Jun 23 08:58 home-assistant_v2.db-wal
And it does not seem to store anything. If this is because the recorder is or is-not working I do not know. All I get is present live data. No logging, or no access to logging, nor history
>> Any way to recover from this ?? The years of history are worth a bunch ....
And if NOT:
>> restart from scratch with new dB, and then restore backup ?
With a 5GB database, that one is definitely not a brand new one.
Stating the obvious, but you won't have history for the past week, i.e. since the backup.
That's a warning you can ignore. It just means the Db was not closed cleanly.
Just notice the current Size, and take a peak before you go to bed, Daily Purge Occur around 4 am ( Not sure the exact time ) , so check in again in the morning ... i guess that's the easiest way, another is using i.e sqlite-browser
( And if you have scripts and templates etc. , requesting datas ( i.e. from past week or 2, they will not find any, and causing errors and warnings )
PS: Default Purge is 10 Days, and as ChrisB says those 2-3 week "statistics" are gone and start from scratch ( or built up upon the 2-3 week old LT-DB-Datas "leaving a gap" )
I believe you, thou i have also no idea here, as you don't show anything, im sure you are not just telling a story, or
OOH Wait , i just notice
Does that means when you try to access System/Settings/Logs it's just shows a "Spinning thingy" ?
In that case i had same issue few releases ago , i believed it was do to me not following my "procedures" as usual ...i.e checking logs before Updating, and fix and perform a restart, before upgrading ( Lesson Learned )
So before you go any further, post your About# System Ver. screenshot ( Also from which core etc You upgraded from )
And check your Core-Log, prior to your first update attempt
Your Lesson would be ( Backup Before Update ) Then you'll only "loose" a few hours of datas
I must admit (with embarrassment), that I somehow put an "include" statement for the recorder in the configuration.yaml, with very few sensors.
As I was having trouble with HA restarts, I had not done any restart for a month, and somehow "forgot" what I had done... thus when there was a restart, the result was no Recording on 99.9 % of the variables. Removing the "Include" statement returned everything normal.
The problem was entirely my own doing... Thanks for your willingness to help...
Bernard
PS: As I am not good at databases, and equally did not want to loose the history of the past years, I sought the help of ChatGPT. Must say quite enlightening. Totally assisted this NOOB to troubleshoot and narrow down the issue...
, i believed it was do to me not following my "procedures" as usual ...i.e checking logs before Updating, and fix and perform a restart, before upgrading ( Lesson Learned )