Skip to main content

Migrating from Imply Enterprise Hybrid to Polaris

If you're currently running Imply Enterprise Hybrid you may be thinking about migrating to Imply Polaris - a fully managed Imply stack in the cloud. You can read about the benefits of Polaris here.

Our migration team is skilled and experienced in enabling customers using a Hybrid architecture to move to Polaris and start enjoying the benefits of a fully managed solution. We will plan your migration with you and then perform as much of the work as we can for you - to get you up and running on Polaris as quickly and smoothly as possible.

If you're interested in migrating from Hybrid to Polaris, contact your Imply account representative.

Differences between Hybrid and Polaris

There are some differences in functionality between the fully managed Polaris service and Imply Enterprise Hybrid.

Infrastructure

If you're used to managing Apache Druid servers in Imply Manager or Helm charts in a Kubernetes deployment, you don't need to manage these things in Polaris. Instead of being concerned about and tuning infrastructure, you can focus on your analytics applications.

Security and user access

We will attempt to set up a method of access to Polaris that is similar to your setup for Hybrid. Our exact approach will depend on the interfaces you're using to control access to Druid.

Querying and ingestion

We will look at the ways in which your users and systems are sending queries to Druid. Polaris will host your Pivot visualizations, and queries sent to a Druid router or broker will be handled by the Polaris API.

Polaris uses familiar Druid ingestion methods, with the additional support included in a fully hosted solution. We will migrate your batch and streaming ingestion methods into Polaris, and we can migrate historical segments if required.

Visualizations

In Polaris you can only use Pivot 2 data cubes and dashboards, not Pivot classic. We will help you assess the feasibility of migrating your classic Pivot resources to Pivot 2.

Administration and monitoring

If you’re used to monitoring Druid with Clarity, you’ll notice that the Polaris monitoring UI and API don’t offer all the metrics you might be familiar with. Generally, the missing metrics are concerned with issues that Imply is responsible for monitoring and fixing in Polaris.

See Polaris limitations for a list of other factors to consider.

Pre-migration

Before the migration, you might want to clean up and delete any unwanted Pivot resources: data cubes, dashboards, reports, users, and roles.

Prior to the migration, Imply will:

  • Work with you to upgrade the Hybrid cluster to the most recent version if possible. If this isn't possible, we will need to determine whether or not we can migrate you directly to Polaris from your existing version.
  • Work with you to enable our access to your Hybrid environment to perform the migration.

Note that in most cases, you will need to point your data source to both Hybrid and Polaris for a period of time, so that data flows in both directions. This will enable you to test in Polaris before cutover. The previous environment can serve as a rollback option to lower the risk of migration - eventually with your agreement we will decommission it.

Migration

During the migration, Imply will do the following:

  • Migrate Pivot users, roles, and groups from Hybrid to Polaris and map Pivot roles to Polaris roles. In Hybrid your users exist in different places—Pivot, Imply Manager. In Polaris your users are in one place.
  • If you have your own LDAP or OIDC solution we will work with you to configure Polaris with the same solution.
  • Migrate event data from the Imply distribution of Apache Druid to Polaris.
  • Migrate classic Pivot resources to Pivot 2.
  • Migrate Pivot resources including data cubes, dashboards, alerts, and reports.
  • Set up resource-based access control by syncing Polaris user IDs with Pivot resources.

Post-migration

Once the migration is complete, we will ensure that your Polaris configuration contains the data and configuration you expect. For example, we will ensure that:

  • Users can log in and can access the same resources as in Hybrid.
  • Queries on data cubes and dashboards work the same as before.
  • Alerts and reports function as expected.

Next steps

Once you're up and running on Polaris, check out the following tutorials and guides. They will help you to adjust to the operational differences between Hybrid and Polaris: