homebridge with big fakegato history - start fails over 90 times, until set fakegato to false and back to true

[nicoh88]

Hello @simont77,

i have a very strange issue with fakegato-history. 😢
I implemented since a long time in homebridge-devolo.

I've had the problem many times when I was still using a Raspberry Pi. But after 5-10 minutes homebridge started normally and without errors. Back then I was thinking of Read / Write IOs? Or the fact that status changes (power consumption, motion detector, contact sensor) were tried to write while loading the faktegato-history. So I then added the isHistoryLoaded function to onCurrentValueChanged.

Whatever, after I moved my homebridge installation to an Intel i3 with SSD, I never had any problems again. Til today ...

---

I try to summarize it.

Homebridge tried 93 times ... until I intervened manually.

Mar 12 13:11:04 frida systemd[1]: homebridge1.service: Scheduled restart job, restart counter is at 93.

The debug log always looks similar, without a real error message the start process fails. I cannot send you the entire log with all 93 attempts - because it exceeds 220 MB. I attached the last attempts here as TXT (2.3 MB).

fakegato-01-true_failstart-over-90-times_cut.txt

Now it's getting interesting. I then deactivated fakegato for homebridge-devolo in config.json and homebridge started properly. (Logfile size 43 KB)

fakegato-02-false_now-starts-fine.txt

Then, I finished homebridge clean again, activated fakegato. Simsalabim! Homebridge now starts properly with activate fakegato. (Logfile 2.3 MB) Many many historical data.

fakegato-03-true-again-starts-fine.txt

---

Phew. Strange right? @simont77

Do you have any idea?

My thoughts:

• Disable print all historical data?
• Homebridge has obviously not ended "properly". Did fakegato-history attempt to write any cached / complete data at startup before isHistoryLoaded? Which was rejected after fakegato: false?
• ❓

Thanks for your time.
Nico