Autoscale for Azure HDInsight is now in preview
Published date: 06 May, 2019
The auto-scale capability for Azure HDInsight is now available in preview. Autoscale automatically scales your Spark, Hive or MapReduce HDInsight clusters up or down based on load or a pre-defined schedule.
- Load-based auto-scale uses several workload-specific metrics, such as CPU and memory usage, to intelligently scale your cluster between user-configurable minimum and maximum sizes based on the load.
- Schedule-based auto-scale allows you to set your own custom schedule for cluster scaling. For example, you can set your cluster to scale up to 10 nodes starting at 9:00 AM and scale back down to 3 nodes at 9:00 PM on work days.