City level geolocation causing Plausible v1.5.1 to crash on startup

[perfectly-preserved-pie]

Hey, I'm following the instructions here: #2324 (comment)

But it's crashing with:

plausible-plausible-1            | {"Kernel pid terminated",application_controller,"{application_start_failure,plausible,{bad_return,{{'Elixir.Plausible.Application',start,[normal,[]]},{'EXIT',{timeout,{'Elixir.Task',await,[#{'__struct__' => 'Elixir.Task',owner => <0.2926.0>,pid => <0.2930.0>,ref => #Ref<0.451273697.225771521.164627>},30000]}}}}}}"}
plausible-plausible-1            | Kernel pid terminated (application_controller) ({application_start_failure,plausible,{bad_return,{{'Elixir.Plausible.Application',start,[normal,[]]},{'EXIT',{timeout,{'Elixir.Task',await,[#{'__struct__' => 'Elixir.Task',owner => <0.2926.0>,pid => <0.2930.0>,ref => #Ref<0.451273697.225771521.164627>},30000]}}}}}})

In my plausible-conf.env file, I have

IP_GEOLOCATION_DB=/etc/app/cities/geolite2-city.mmdb
GEONAMES_SOURCE_FILE=/etc/app/cities/geonames.csv

In my docker-compose.yml file, I have:

plausible:
    hostname: plausible
    image: plausible/analytics:latest
    restart: always
    command: sh -c "sleep 10 && /entrypoint.sh db createdb && /entrypoint.sh db migrate && /entrypoint.sh db init-admin && /entrypoint.sh run"
    depends_on:
      - plausible_db
      - plausible_events_db
      - mail
    volumes:
      - /appdata/plausible/cities/geolite2-city.mmdb:/etc/app/cities/geolite2-city.mmdb:ro
      - /appdata/plausible/cities/geonames.csv:/etc/app/cities/geonames.csv:ro
    env_file:
      - plausible-conf.env

My filesystem structure on the host:

redacted:/appdata/plausible/cities$ ls -al
total 187912
drwxr-xr-x 2 root root      4096 Feb 21 12:17 .
drwxr-xr-x 5 root root      4096 Jan  5 20:25 ..
-rw-r--r-- 1 root root  70799360 Feb 21 12:17 geolite2-city.mmdb
-rw-r--r-- 1 root root 121603380 Feb 21 11:38 geonames.csv

If I comment out the GEONAMES_SOURCE_FILE in the .env file, Plausible works again.

Any idea where I went wrong?

[ruslandoga]

👋 @perfectly-preserved-pie

This error means Plausible times out when loading some geoname_id - city name mappings. One possible workaround is to rebuild the image with a higher timeout value or use a smaller geonames set (e.g. the default).

More info: #2324 (reply in thread)

[perfectly-preserved-pie]

👋 @perfectly-preserved-pie

This error means Plausible times out when loading some geoname_id - city name mappings. One possible workaround is to rebuild the image with a higher timeout value or a smaller geonames set.

More info: #2324 (reply in thread)

Interesting, thanks! I completely missed that linked reply in the original thread... whoops.

Is there a big difference between the built-in geonames and the one provided in that thread? If not, and as long as I can still get decently accurate city data, I think I'll just continue to use the built-in geonames.

[ruslandoga]

According to :location, the built-in geonames mapping is based on cities500.txt which are "all cities with a population > 500 or seats of adm div down to PPLA4 (ca 185.000)"

[perfectly-preserved-pie]

Very cool. I appreciate the detailed explanation. I'll stick with the default mapping.