Downscale, upscale, and rebalance clusters automatically in the cloud based on SLA, priority, and workload context of each job.


Prevent cost overruns by shutting down idle nodes upon job completion.

Use Aggressive Downscaling to rebalance workloads across active nodes and decommission idle ones without the risk of data loss. Enable faster recycling of clusters and nodes while simultaneously providing cost savings, stability, performance, and fault tolerance benefits.


Get additional utilization of the existing compute nodes instead of adding additional nodes.

Optimized upscaling avoids wasted/underutilized resources by recapturing them and helps with greater cost avoidance.

Data Allocation

Workload packing performs smart allocation of workloads, freeing up larger pools of nodes to downscale while preventing cluster hot spots and honoring data locality preferences.

This novel non-uniform resource allocation strategy further reduces the cost to run elastic clusters.

As we made the transition to the cloud, Qubole's ability to automate the infrastructure and easily scale to meet the demands of our users saved us time and resources, and reduced our TCO by over $700k. - Wade Warren, SVP Global Engineering and Tech Ops, Wikia
We’ve been able to lower our Presto compute costs quite a lot with Qubole, because we can depend on Qubole’s autoscaling and automated Spot buying to adjust our cluster size quickly and appropriately." - Russell Rhodes, Senior Manager, Big Data and head of Data Infrastructure Team, Zillow