Merge pull request #61 from sebastianrakel/fix-postgres-online-check

Don't use wait for host port instead use pg_isready

Author: bastelfreak

openvoxdb/Containerfile

@@ -16,7 +16,7 @@ LABEL org.label-schema.maintainer="Voxpupuli Release Team <voxpupuli@groups.io>"
       org.label-schema.build-date="$build_date"
 
 # Use netcat-openbsd to support IPv6 connectivity checks (netcat-traditional is IPv4 only)
-ARG PACKAGES="ca-certificates curl dnsutils netcat-openbsd dumb-init"
+ARG PACKAGES="ca-certificates curl dumb-init postgresql-client postgresql-client-common"
 
 ARG LOGDIR
 ENV LOGDIR=${LOGDIR:-/opt/puppetlabs/server/data/puppetdb/logs}

openvoxdb/container-entrypoint.d/10-wait-for-hosts.sh

@@ -16,7 +16,6 @@ set -e
 #   OPENVOXDB_POSTGRES_HOSTNAME       Specified in Dockerfile, defaults to postgres
 #   OPENVOXSERVER_HOSTNAME            DNS name of puppetserver to wait on, defaults to puppet
 
-
 msg() {
     echo "($0) $1"
 }
@@ -26,43 +25,18 @@ error() {
     exit 1
 }
 
-# Alpine as high as 3.9 seems to have failures reaching addresses sporadically
-# In local repro scenarios, performing a DNS lookup with dig increases reliability
-wait_for_host_name_resolution() {
-  # host and dig are in the bind-tools Alpine package
-  # k8s nodes may not be reachable with a ping
-  # performing a dig prior to a host may help prime the cache in Alpine
-  # https://github.com/Microsoft/opengcs/issues/303
-  /wtfc.sh --timeout="${2}" --interval=1 --progress "dig $1 && host $1"
-  # additionally log the DNS lookup information for diagnostic purposes
-  NAME_RESOLVED=$?
-  dig $1
-  if [ $NAME_RESOLVED -ne 0 ]; then
-    error "dependent service at $1 cannot be resolved or contacted"
-  fi
-}
-
-wait_for_host_port() {
-  # -v verbose -w connect / final net read timeout -z scan and don't send data
-  /wtfc.sh --timeout=${3} --interval=1 --progress "nc -v -w 1 -z '${1}' ${2}"
-  if [ $? -ne 0 ]; then
-    error "host $1:$2 does not appear to be listening"
-  fi
-}
-
 OPENVOXDB_WAITFORHOST_SECONDS=${OPENVOXDB_WAITFORHOST_SECONDS:-30}
 OPENVOXDB_WAITFORPOSTGRES_SECONDS=${OPENVOXDB_WAITFORPOSTGRES_SECONDS:-60}
 OPENVOXDB_WAITFORHEALTH_SECONDS=${OPENVOXDB_WAITFORHEALTH_SECONDS:-360}
 OPENVOXDB_POSTGRES_HOSTNAME="${OPENVOXDB_POSTGRES_HOSTNAME:-postgres}"
 OPENVOXSERVER_HOSTNAME="${OPENVOXSERVER_HOSTNAME:-puppet}"
 OPENVOXSERVER_PORT="${OPENVOXSERVER_PORT:-8140}"
 
-# wait for postgres DNS
-wait_for_host_name_resolution $OPENVOXDB_POSTGRES_HOSTNAME $OPENVOXDB_WAITFORHOST_SECONDS
+# wait for postgres is ready
+/wtfc.sh --timeout="${OPENVOXDB_WAITFORHOST_SECONDS}" --interval=1 --progress "pg_isready -h ${OPENVOXDB_POSTGRES_HOSTNAME} --port '${OPENVOXDB_POSTGRES_PORT:-5432}'"
 
 # wait for puppetserver DNS, then healthcheck
 if [ "$USE_OPENVOXSERVER" = true ]; then
-  wait_for_host_name_resolution $OPENVOXSERVER_HOSTNAME $OPENVOXDB_WAITFORHOST_SECONDS
   HEALTH_COMMAND="curl --silent --fail --insecure 'https://${OPENVOXSERVER_HOSTNAME}:"${OPENVOXSERVER_PORT}"/status/v1/simple' | grep -q '^running$'"
 fi
 
@@ -72,6 +46,3 @@ if [ -n "$HEALTH_COMMAND" ]; then
     error "Required health check failed"
   fi
 fi
-
-# wait for postgres
-wait_for_host_port $OPENVOXDB_POSTGRES_HOSTNAME "${OPENVOXDB_POSTGRES_PORT:-5432}" $OPENVOXDB_WAITFORPOSTGRES_SECONDS