#!/bin/bash

DELAY="{{ handler_health_osd_check_delay }}"
CEPH_CLI="--name client.bootstrap-osd --keyring /var/lib/ceph/bootstrap-osd/{{ cluster }}.keyring --cluster {{ cluster }}"

check_pgs() {
  num_pgs=$($docker_exec ceph $CEPH_CLI -s -f json|python -c 'import sys, json; print(json.load(sys.stdin)["pgmap"]["num_pgs"])')
  if [[ "$num_pgs" == "0" ]]; then
    return 0
  fi
  while [ $RETRIES -ne 0 ]; do
    test "$($docker_exec ceph $CEPH_CLI -s -f json | python -c 'import sys, json; print(json.load(sys.stdin)["pgmap"]["num_pgs"])')" -eq "$($docker_exec ceph $CEPH_CLI -s -f json | python -c 'import sys, json; print sum ( [ i["count"] for i in json.load(sys.stdin)["pgmap"]["pgs_by_state"] if "active+clean" in i["state_name"]])')"
    RET=$?
    test $RET -eq 0 && return 0
    sleep $DELAY
    let RETRIES=RETRIES-1
  done
  # PGs not clean, exiting with return code 1
  echo "Error while running 'ceph $CEPH_CLI -s', PGs were not reported as active+clean"
  echo "It is possible that the cluster has less OSDs than the replica configuration"
  echo "Will refuse to continue"
  $docker_exec ceph $CEPH_CLI -s
  $docker_exec ceph $CEPH_CLI osd dump
  $docker_exec ceph $CEPH_CLI osd tree
  $docker_exec ceph $CEPH_CLI osd crush rule dump
  exit 1
}

wait_for_socket_in_docker() {
  osd_mount_point=$(docker exec "$1" df --output=target | grep '/var/lib/ceph/osd/')
  whoami=$(docker exec "$1" cat $osd_mount_point/whoami)
  if ! docker exec "$1" timeout 10 bash -c "while [ ! -e /var/run/ceph/ceph-osd.${whoami}.asok ]; do sleep 1 ; done"; then
    echo "Timed out while trying to look for a Ceph OSD socket."
    echo "Abort mission!"
    exit 1
  fi
}

get_docker_id_from_dev_name() {
  local id
  local count
  count=10
  while [ $count -ne 0 ]; do
    id=$(docker ps -q -f "name=${1}$")
    test "$id" != "" && break
    sleep $DELAY
    let count=count-1
  done
  echo "$id"
}

# The unit file looks like: ceph-osd@NNN.service where NNN is OSD ID
for unit in $(systemctl list-units | grep -E "loaded * active" | grep -oE "ceph-osd@[0-9]+.service"); do
  # First, restart daemon(s)
  systemctl restart "${unit}"
  # We need to wait because it may take some time for the socket to actually exists
  COUNT=10
  # Wait and ensure the socket exists after restarting the daemon
  osd_id=$(echo ${unit#ceph-osd@} | grep -oE '[0-9]+')
  {% if containerized_deployment -%}
  container_id=$(get_docker_id_from_dev_name "ceph-osd-${osd_id}")
  docker_exec="docker exec $container_id"
  {% endif %}
  SOCKET=/var/run/ceph/{{ cluster }}-osd.${osd_id}.asok
  while [ $COUNT -ne 0 ]; do
    RETRIES="{{ handler_health_osd_check_retries }}"
    $docker_exec test -S "$SOCKET" && check_pgs && continue 2
    sleep $DELAY
    let COUNT=COUNT-1
  done
  # If we reach this point, it means the socket is not present.
  echo "Socket file ${SOCKET} could not be found, which means the osd daemon is not running. Showing ceph-osd unit logs now:"
  journalctl -u "${unit}"
  exit 1
done