Start/stop/restart a cluster

Dynatrace Managed software consists of a number of Dynatrace services that are dependent on each other and should be stopped or started in a particular order. Use this procedure for Dynatrace Managed deployments containing three or more nodes.

For Dynatrace Managed deployments containing less than three (3) nodes, you can use the official dynatrace.sh script with additional parameters to properly start, stop, or restart Dynatrace services on each node individually within the cluster (Start/stop/restart a node).
For Dynatrace Managed deployments containing three (3) or more nodes, use the following cluster procedures:

Shut down cluster

Issue the following command on each existing node:

sudo nohup ./opt/dynatrace-managed/launcher/server.sh stop

Issue the following command to check the status:
```
sudo nohup ./opt/dynatrace-managed/launcher/server.sh status
```
When all server processes are fully shut down and a status check shows them as not running, issue the following command on each existing node, one node at a time:
```
sudo nohup ../launcher/dynatrace.sh stop
```

Start up cluster

Run /opt/dynatrace-managed/launcher/firewall.sh start on each node.
Issue the following command on each existing node (wait a minute between each startup):
```
sudo nohup ./opt/dynatrace-managed/launcher/elasticsearch.sh start
```
Issue the following command to check the status:
```
curl -X GET "localhost:9200/_cluster/health?wait_for_status=yellow&timeout=50s&pretty"
```
Check status
Make sure active_shards_percent_as_number is 100% and number_of_nodes is equal to the number of nodes in the cluster.
Issue the following command on each existing node (wait a minute between each startup):
```
sudo nohup ./opt/dynatrace-managed/launcher/cassandra.sh start
```
Issue the following command to check the status:
```
sudo nohup ./opt/dynatrace-managed/utils/cassandra-nodetool.sh status
```
Check status
Make sure that all nodes display UN before proceeding to the next step.
Issue the following command on each existing node. (Wait a minute between each startup. It can take up to ~10 minutes to fully start a node.)
```
sudo nohup ./opt/dynatrace-managed/launcher/server.sh start
```
Once all server processes are up and running, start the rest of the processes. Issue the following command on each existing node:
```
sudo nohup ./opt/dynatrace-managed/launcher/dynatrace.sh start
```