Remote log monitors over ssh must reconnect: wrap 'ssh host tail -F' in a retry loop

A watcher like ssh host 'tail -n0 -F log' | grep --line-buffered ... dies for good when the VPN or network drops (ssh exits 255: 'Broken pipe' / 'Operation timed out'). Wrap it in while true; do ssh -o ConnectTimeout=60 -o ServerAliveInterval=60 -o BatchMode=yes host 'tail -n0 -F log' | grep ...; echo once 'connection lost'; sleep 60; done. ServerAliveInterval makes a dead link fail fast instead of hanging. Note tail -n0 drops lines written during the outage; if those matter, track the last line number and resume with tail -n +N.