FROM node:16.14.2-alpine RUN apk add --no-cache dnsmasq=2.86-r1 WORKDIR /usr/app ENV PATH="/usr/app/bin:${PATH}" # schedule critical checks to run every 5 minutes (any failures will disable server) # schedule extended checks to run on every hour (optional checks, report only) RUN echo '*/5 * * * * source /etc/environment ; /usr/app/bin/cli run critical >> /proc/1/fd/1' >> /etc/crontabs/root && \ echo '0 * * * * source /etc/environment ; /usr/app/bin/cli run extended >> /proc/1/fd/1' >> /etc/crontabs/root COPY packages/health-check/package.json \ packages/health-check/yarn.lock \ ./ RUN yarn --frozen-lockfile COPY packages/health-check/src src COPY packages/health-check/cli cli COPY packages/health-check/bin bin EXPOSE 3100 ENV NODE_ENV production # 1. get public server ip and save it in /etc/environment (passed to cron tasks as env variable) # 2. start dnsmasq in the background with: # - alias PORTAL_DOMAIN with current server ip so it overrides potential load balancer request # - default docker nameserver 127.0.0.11 for any other request # 3. replace docker nameserver with dnsmasq nameserver in /etc/resolv.conf # 4. start crond in the background to schedule periodic health checks # 5. start the health-check api service CMD [ "sh", "-c", \ "export serverip=$(node src/whatismyip.js) && \ echo \"export serverip=${serverip}\" >> /etc/environment && \ dnsmasq --no-resolv --log-facility=/var/log/dnsmasq.log --address=/$PORTAL_DOMAIN/$serverip --server=127.0.0.11 && \ echo \"$(sed 's/127.0.0.11/127.0.0.1/' /etc/resolv.conf)\" > /etc/resolv.conf && \ crond && \ node src/index.js" \ ]