telegraf log wrangling to scrape apache or nginx logs for erddap related requests and sent to plausible or google analytics event api's see https://www.influxdata.com/time-series-platform/telegraf/ for more details on teleraf
create env file
cp .env.sample .env
edit .env values. This to check are
- LOG_FILE - this should point to the nginx or apache log folder. not directly to a file.
- HOST_URL - hostname to report in plausible with protocol
- DOMAINS - keys used by plausible to route events see https://plausible.io/docs/events-api#request-body-json-parameters
- PLAUSIBLE_URL - url to plausible instance
- SENTRY_URL - url to sentry project for monitering up time
review telegraf/telegraf.conf. the most important one being
filesunder [[inputs.tail]]
find the group that has read access to the log files
stat -c '%g' /var/log/apache2
use the resulting number to set the group in the telegraf container by editing the docker compose and adding
user: telegraf:4
where 4 is the gid of the group
Start the container
sudo docker-compose up -d
- log file name is hard coded i.e. catalogue-ssl-access.log. it would be nice to make it based on an environment variable.