Monitor clickhouse proxy for connection issues

Right now the Clickhouse proxy is a tier 0 component, all Clickhouse traffic in AWS must go through the Clickhouse proxy. If this component fails, every service in AWS using Clickhouse will fail.

One possible failure could be that the Clickhouse proxy is redeployed and the firewall rules in data1 are not updated to match the new ip address of the redeployed server

We need a good monitoring to detect issues with the clickhouse proxy as swift as possible, particularly with the connection between Clickhouse and the proxy

A possible solution could be:

1. Add a timer to the clickhouse proxy to ping the database once every couple minutes
2. Log the result of the ping to grafana
3. Add alerts to grafana for:
   4. Failed pings: The server did not respond or responded with an error
   5. Missing pings: The server is not sending ping status, meaning that maybe the entire machine is down

Note that this becomes increasingly important as we move more components from the monolith to AWS. See: https://github.com/ooni/backend/issues/807#issuecomment-3103559446