Health

Development

curl -fsS http://127.0.0.1:8091/health
curl -fsS http://127.0.0.1:8091/institution/status
curl -fsS http://127.0.0.1:8093/healthz
curl -fsS http://127.0.0.1:8094/health
# If the conversation-ui stack is up:
curl -s -o /dev/null -w '%{http_code}\n' http://127.0.0.1:3080/
containers ps --format '{{.Names}}\t{{.Status}}' | grep sis-conversation-ui
ci-check          # compose + suites + guardians; does not dump secrets

ci-check is the single validation. Do not run compose config on its own: it prints secrets.

VPS

containers ps -a --format '{{.Names}}\t{{.Status}}' | grep -E 'society|sis-conversation-ui'
ss -ltnp | grep -E '8091|8093|8094|3080|3081'
curl -s -o /dev/null -w '%{http_code}\n' https://sis.<client>.do/

tools/ops/prod-healthcheck.sh every 5 min → README badge (healthchecks.io). Red = some VPS stopped reporting, not a fine diagnosis.

Orphan sweep (prod-up.sh also does this at the end):

for c in $(containers ps --format '{{.Names}}' | grep '^society'); do
  net=$(containers inspect $c --format '{{range $k,$v := .NetworkSettings.Networks}}{{$k}}{{end}}')
  [ -z "$net" ] && echo "$c NO NETWORK"
done

edge proxy has no log directive on the reference client proxy: no access log. Real traffic = upstream.

After touching edge proxy: the usual 11 domains are the canary (including onto.* when it exists). A 200 on the new service and a 502 on chat. is a half deploy.