Pepperdata's Secret Sauce

Pepperdata Capacity Optimizer provides the system scheduler with real-time visibility into actual hardware utilization so pending pods can be launched on existing nodes with allocated but unused capacity—resulting in more pods per node, greater utilization, and lower cost.

Capacity Optimizer also enhances the efficiency of your cloud autoscaler by ensuring new nodes are provisioned only when existing nodes are fully utilized.

If there are any pods in the pending state, Pepperdata Capacity Optimizer enables the autoscaler (e.g. Karpenter) to add new nodes only when all existing nodes are fully utilized.

As soon as you enable Pepperdata optimization, you’ll immediately start to see measurable resource utilization uplift to 80 percent or more—not days or weeks later.

Your workloads and your cluster environment are dynamic, making it difficult, if not impossible, to predict peak traffic times. Other solutions must be manually trained on typical workload patterns, which can cause the solution to break if those patterns change.

Pepperdata provides the system scheduler with a real-time stream of data so scheduling decisions are based on actual resource usage instead of allocations.

Pepperdata requires no manual intervention and operates on your behalf to automatically eliminate waste as your applications run.

Pepperdata works around the clock automatically and never stops optimizing your workloads.

Pepperdata software is safe to deploy because it never touches or changes your running applications. Our optimization is “node aware,” meaning it maintains your nodes at their optimal “sweet spot” at all times.

Pepperdata optimization ensures both:

1. Optimal capacity utilization by providing the system scheduler with real-time visibility into actual hardware utilization so pending pods can be launched on existing nodes with allocated but unused capacity
2. Optimal cost by ensuring that every cluster or node is fully utilized before a new node is added

No other solution in the market optimizes your clusters in both these ways.

Pepperdata supports optimization of these Kubernetes and YARN workloads running on-premises or in the cloud:

• Apache Spark
• Apache Flink
• Apache Airflow
• Jobs and CronJobs
• Custom workloads based on labels and annotations
• Microservices

Pepperdata frees your development teams from the tedium of tweaking and tuning applications and applying recommendations so they can focus on high-value, innovative activities to grow your business.

Even after you’ve deployed other optimization tools, Pepperdata still saves you an additional 30 percent on average.

Pepperdata works side-by-side with optimization solutions such as Karpenter and instance rightsizing to recapture overprovisioning waste from your workloads for increased utilization, improved throughput performance, and reduced cost.