Announcing Imply 3.2

by Vadim Ogievetsky · January 30, 2020

We are delighted to announce that Imply 3.2 is now available! Imply 3.2 is based on Apache Druid 0.17 (Druid’s first Apache top level project release) and adds new cloud management, alerting, reporting, and data loading features.

Top level Druid

Imply 3.2 is based on Druid 0.17.0, which you can read about in a separate blog post. The Druid 0.17.0 release includes over 250 new features, performance enhancements, bug fixes, and major documentation improvements.

Highlights include:

  • Load batch data from HDFS without using Hadoop
  • Parallel query merging on brokers for increased performance
  • SQL-compatible null handling for better SQL interconnection with other tools
  • Parallel auto-compaction for better compaction throughput
  • Numerous new readiness endpoints for better interplay with Kubernetes

Cluster management in your cloud

This release includes a fully managed solution for your cloud. The cloud manager lets you deploy and manage clusters in your own cloud via Kubernetes.

Newly released cloud management functionality includes:

  • Point-and-click control of the Druid cluster
  • One-click cluster scale up
  • Central control of versions and upgrades

Scheduled reports

It is now possible to define regularly scheduled reports to be delivered in-application or via email. Scheduled report configurations support all existing data export formats, such as CSV, JSON, and Excel as email attachments. Additionally, scheduled reports support email delivery to external, non-Pivot users.

Alerts GA

This release brings many improvements to Imply Pivot’s alerting functionality. Alerts can be configured against multiple condition thresholds, checking against either aggregate measures for a dimension or against dimension values, and users will be notified when alerts are triggered via either email or webhook.

Native reading from HDFS

In this release we are adding the ability for Druid to read binary data (e.g. ORC, Parquet) from HDFS directly. This means that if your data is stored in a data lake you can now index it in Druid via a point and click wizard directly, without using Hadoop.

Alerts in Imply Clarity

Alerts will also be made available in Imply Clarity, allowing you to set up alerting for the health of your Druid cluster. These alerts could proactively notify you via email or a webhook integration (such as PagerDuty) about any issues in your cluster according to your own SLAs and limits.

Customization improvements

This release also expands Pivot’s customization ability. Pivot can now be configured to hide information about the underlying version and can also be configured to display a custom message when a logged-in user does not have any permissions, e.g. when a user authorized via LDAP has no matching groups.

30-Day Free Trial

The Imply 3.2 distribution is now available for download as a 30-day trial, or you can subscribe to a 30-day trial of the Imply Cloud managed service based on Imply 3.2, at https://imply.io/get-started.

Back to blog

How can we help?