openstack-helm-infra/elasticsearch/templates/bin/_elasticsearch.sh.tpl
willxz e89c1c3c06 allocate_data_node function improvement
- Remove "if" condition of allocate_data_node
- Dealy 5 seconds for wait_to_join initial check to start
- Set 60 minutes timeout for wait_to_join function

Change-Id: Ie42af89551bd8804b87fe936c676e85130564187
2020-07-07 12:28:21 -04:00

131 lines
4.8 KiB
Smarty

#!/bin/bash
{{/*
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/}}
{{- $envAll := . }}
set -e
COMMAND="${@:-start}"
function initiate_keystore () {
bin/elasticsearch-keystore create
echo ${S3_ACCESS_KEY} | /usr/share/elasticsearch/bin/elasticsearch-keystore add -xf s3.client.default.access_key
echo ${S3_SECRET_KEY} | /usr/share/elasticsearch/bin/elasticsearch-keystore add -xf s3.client.default.secret_key
}
function start () {
ulimit -l unlimited
initiate_keystore
exec /usr/local/bin/docker-entrypoint.sh elasticsearch
}
function stop () {
kill -TERM 1
}
function wait_to_join() {
# delay 5 seconds before the first check
sleep 5
joined=$(curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" "${ELASTICSEARCH_ENDPOINT}/_cat/nodes" | grep -w $NODE_NAME || true )
i=0
while [ -z "$joined" ]; do
sleep 5
joined=$(curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" "${ELASTICSEARCH_ENDPOINT}/_cat/nodes" | grep -w $NODE_NAME || true )
i=$((i+1))
# Waiting for up to 60 minutes
if [ $i -gt 720 ]; then
break
fi
done
}
function allocate_data_node () {
echo "Node ${NODE_NAME} has started. Waiting to rejoin the cluster."
wait_to_join
echo "Re-enabling Replica Shard Allocation"
curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" -XPUT -H 'Content-Type: application/json' \
"${ELASTICSEARCH_ENDPOINT}/_cluster/settings" -d "{
\"persistent\": {
\"cluster.routing.allocation.enable\": null
}
}"
}
function start_master_node () {
ulimit -l unlimited
initiate_keystore
if [ ! -f {{ $envAll.Values.conf.elasticsearch.config.path.data }}/cluster-bootstrap.txt ];
then
{{ if empty $envAll.Values.conf.elasticsearch.config.cluster.initial_master_nodes -}}
{{- $_ := set $envAll.Values "__eligible_masters" ( list ) }}
{{- range $podInt := until ( atoi (print $envAll.Values.pod.replicas.master ) ) }}
{{- $eligibleMaster := printf "elasticsearch-master-%s" (toString $podInt) }}
{{- $__eligible_masters := append $envAll.Values.__eligible_masters $eligibleMaster }}
{{- $_ := set $envAll.Values "__eligible_masters" $__eligible_masters }}
{{- end -}}
{{- $masters := include "helm-toolkit.utils.joinListWithComma" $envAll.Values.__eligible_masters -}}
echo {{$masters}} >> {{ $envAll.Values.conf.elasticsearch.config.path.data }}/cluster-bootstrap.txt
exec /usr/local/bin/docker-entrypoint.sh elasticsearch -Ecluster.initial_master_nodes={{$masters}}
{{- end }}
else
exec /usr/local/bin/docker-entrypoint.sh elasticsearch
fi
}
function start_data_node () {
ulimit -l unlimited
initiate_keystore
allocate_data_node &
/usr/local/bin/docker-entrypoint.sh elasticsearch &
function drain_data_node () {
# Implement the Rolling Restart Protocol Described Here:
# https://www.elastic.co/guide/en/elasticsearch/reference/7.x/restart-cluster.html#restart-cluster-rolling
echo "Disabling Replica Shard Allocation"
curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" -XPUT -H 'Content-Type: application/json' \
"${ELASTICSEARCH_ENDPOINT}/_cluster/settings" -d "{
\"persistent\": {
\"cluster.routing.allocation.enable\": \"primaries\"
}
}"
# If version < 7.6 use _flush/synced; otherwise use _flush
# https://www.elastic.co/guide/en/elasticsearch/reference/current/indices-synced-flush-api.html#indices-synced-flush-api
version=$(curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" "${ELASTICSEARCH_ENDPOINT}/" | jq -r .version.number)
if [[ $version =~ "7.1" ]]; then
action="_flush/synced"
else
action="_flush"
fi
curl -s -K- <<< "--user ${ELASTICSEARCH_USERNAME}:${ELASTICSEARCH_PASSWORD}" -XPOST "${ELASTICSEARCH_ENDPOINT}/$action"
# TODO: Check the response of synced flush operations to make sure there are no failures.
# Synced flush operations that fail due to pending indexing operations are listed in the response body,
# although the request itself still returns a 200 OK status. If there are failures, reissue the request.
# (The only side effect of not doing so is slower start up times. See flush documentation linked above)
echo "Node ${NODE_NAME} is ready to shutdown"
kill -TERM 1
}
trap drain_data_node TERM EXIT HUP INT
wait
}
$COMMAND