Docker Quickstart (preview)

Imply Manager is currently being offered as part of a preview program. To participate in this program, please contact your Imply representative.

In this quickstart, we will walk through configuring and running the Imply Manager on your local machine. This quickstart assumes that you are using OSX, but it can be easily adapted to support Linux or other Unix-like operating systems.

Prerequisites

To use the Imply Manager, you will need:

• Docker (install). If using Docker Desktop, give at least 2 CPU cores and 4GB of memory to the Docker Engine.
• a Docker ID to access private images on Docker Hub (signup)
• an Imply Manager license key (contact your Imply representative)

Login to Docker Hub

Imply Manager is provided as a pair of Docker images available on Docker Hub:

• the manager (imply/onprem-manager) - runs the coordination services and manager application
• the agent (imply/onprem-agent) - becomes the nodes of one or more Druid clusters

These are private repositories, so you will need to provide your Docker ID to Imply in order to gain access. Once this is done, log in with your Docker ID credentials:

From a Terminal window:

docker login

Set up dependencies

Imply Manager requires ZooKeeper and a MySQL server to operate. In this quickstart, we will run these dependencies as additional Docker containers on your local machine. To ensure that our containers can communicate with each other, we will first create a user-defined bridge network with the following command:

docker network create imply

Start ZooKeeper:

docker run --name zk --network imply -d --rm zookeeper:3.4.14

Start MySQL:

docker run --name mysql --network imply -e MYSQL_ROOT_PASSWORD=imply -d --rm mysql:5.7.26

Write configuration file

Create a file named env.list with the following contents, substituting {LICENSE_KEY} with the license key provided by Imply:

IMPLY_MANAGER_LICENSE_KEY={LICENSE_KEY}

IMPLY_MANAGER_STORE_TYPE=mysql
IMPLY_MANAGER_STORE_HOST=mysql
IMPLY_MANAGER_STORE_PORT=3306
IMPLY_MANAGER_STORE_USER=root
IMPLY_MANAGER_STORE_PASSWORD=imply
IMPLY_MANAGER_STORE_DATABASE=imply_manager

imply_defaults_zkType=external
imply_defaults_zkHosts=zk:2181
imply_defaults_zkBasePath=imply
imply_defaults_metadataStorageType=mysql
imply_defaults_metadataStorageHost=mysql
imply_defaults_metadataStoragePort=3306
imply_defaults_metadataStorageUser=root
imply_defaults_metadataStoragePassword=imply
imply_defaults_deepStorageType=local
imply_defaults_deepStorageBaseLocation=/mnt/var/druid/storage

Run the Imply Manager containers

Start the manager container:

docker run --name imply-manager --network imply -p 9097:9097 --env-file env.list -d --rm imply/onprem-manager

Start an agent container:

docker run --network imply -p 9095:9095 -p 8888:8888 -e "IMPLY_MANAGER_HOST=imply-manager" -d --rm imply/onprem-agent

Creating your first cluster

Once the manager container is running, you can access the Imply Manager in your web browser at: http://localhost:9097

• Create the administrator account and then click Login to proceed.
• From the dashboard, click on + New cluster to create a new cluster definition.
• The default values are all appropriate for this quickstart. Click Create cluster to proceed.
• Click OK to confirm cluster creation.

At this point, the cluster definition has been submitted but no resources have been assigned to the cluster. You may have to wait a few seconds for the cluster to be ready, at which point you should see a link to 'Start by adding a server':

• Click the link to add a new server.
• If the agent was able to connect to the manager container, there should be a single available server listed. Select this server.
• Under Roles, select: dataTier1, master, and query.
• Click OK to close the dialog.
• Click Apply changes and then click Apply with restart on the confirmation dialog.

The Cluster Overview screen should show the cluster in an UPDATING state. The update operation may take a few minutes. Once the update is finished, the cluster's state will change to RUNNING.

Congratulations! Your cluster is now up and running. From here you can visit:

• http://localhost:8888 - to access the new Druid console
• http://localhost:9095 - to access the Pivot application

If you would like to load some example data into your cluster, continue with the Load data file section of the general Quickstart documentation.