Start, Stop, and Status
This section describes how to startup and shutdown Dremio on YARN and non-YARN deployments for RPM and Tarball installations.
Commands for all Deployments
The following commands are used to start up, shut down, and view the status of the nodes in your Dremio cluster.
Usage:
dremio [--config <conf-dir>] (start|start-fg||stop|status|restart|autorestart)
| Command Type | Tarball Command | RPM Command |
|---|---|---|
| Help | <DREMIO-HOME>/bin/dremio |
service dremio |
| Start | <DREMIO-HOME>/bin/dremio start |
service dremio start |
| Stop | <DREMIO-HOME>/bin/dremio stop |
service dremio stop |
| Status | <DREMIO-HOME>/bin/dremio status |
service dremio status |
| Restart | <DREMIO-HOME>/bin/dremio restart |
service dremio restart |
| Autorestart | <DREMIO-HOME>/bin/dremio autorestart |
service dremio autorestart |
YARN Deployments
Use the following instructions to properly shut down or start up a Dremio cluster with a YARN deployment, The commands used for node startup, shutdown, and status is different depending on whether your installation used an RPM or Tarball package.
Starting up with YARN Deployments
To start up a Dremio cluster:
- Manually startup the master-coordinator node first. Use the appropriate startup command (startup command list below) as per your deployment.
- Manually startup the remaining coordinators one-at-a-time waiting for each one to successfully complete before moving on to the next one.
- Navigate to the Admin -> Provisioning screen in the Dremio UI and click the Start Cluster button. This brings up all Dremio executors (running as Yarn containers) in the Hadoop cluster.
Shutting down with YARN Deployments
To shut down a Dremio cluster:
- Navigate to the Admin -> Provisioning screen in the Dremio UI and click the Stop Cluster button. This brings down all Dremio executors (running as Yarn containers) in the Hadoop cluster.
- Manually shutdown all the coordinators (EXCEPT for the master coordinator, if it is a separate node) one-at-a-time waiting for each one to successfully complete before moving on to the next one.
- Manually shutdown the master-coordinator node last.
Viewing Cluster Status with YARN Deployments
To view the status of your Dremio cluster in the Dremio UI, navigate to the Admin -> Node Activity screen.
Non-YARN Deployments
Use the following instructions to properly shut down or start up a Dremio cluster on non-YARN deployments, The commands used for node startup, shutdown, and status is different depending on whether your installation used an RPM or Tarball package.
Starting up with non-YARN Deployments
To start up a Dremio cluster:
- Manually startup the master-coordinator node first. Use the appropriate startup command (startup command list below) as per your deployment.
- Manually startup the remaining coordinators one-at-a-time waiting for each one to successfully complete before moving on to the next one.
- Manually startup all the executors one-at-a-time waiting for each one to successfully complete before moving on to the next one.
Shutting down with non-YARN Deployments
To shut down a Dremio cluster:
- Manually shutdown all the executors one-at-a-time waiting for each one to successfully complete before moving on to the next one.
Use the appropriate shutdown command (shutdown command list below) as per your deployment. - Manually shutdown all the coordinators (EXCEPT for the master coordinator, if it is a separate node) one-at-a-time waiting for each one to successfully complete before moving on to the next one.
- Manually shutdown the master-coordinator node last.