Run More, Spend Less
On Your Kubernetes Workloads—Automatically
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 extra capacity—resulting in more pods per node, greater utilization, and lower workload cost.
-
Improve Resource Utilization
Create more containers per node to increase resource utilization by up to 80%.
-
Reduce Costs Across All Your Clusters
Cut cloud spend by 30% on average with no developer effort needed.
-
Instantly Reclaim Wasted CPU, GPU, and
Memory ResourcesInstantly Reclaim Wasted CPU, GPU, and Memory Resources
Ensure maximum throughput performance by continuously optimizing GPU, CPU, and memory in real time.
Free Engineers from Manual Tuning, Applying Recommendations, or Changing Application Code
Run Pepperdata Capacity Optimizer continuously and automatically in the background with no tedious recommendations to implement and no need for developers to modify code or tune configs.
Pepperdata Capacity Optimizer
Four Components of Intelligent Kubernetes Resource Optimization
-
The Scheduler
Makes more accurate and efficient resource decisions with Pepperdata-provided data.
-
Workloads on Nodes
Are spun up based on real-time physical utilization.
-
Pods
Are launched with optimized resource requests.
-
The Autoscaler
Can scale up more efficiently since nodes are packed based on actual utilization.
Supported Technologies for Pepperdata Capacity Optimizer
Supported Environments
- Apache Spark, Apache Flink, Apache Airflow, Jobs, JobController, and CronJobs on Kubernetes, Amazon EKS, Google GKE
- Apache Spark on Cloudera Data Engineering
- Amazon EMR, Google DataProc
- Cloudera Data Platform (CDP)
Supported Schedulers
- Default scheduler on Amazon EMR and EKS and Google GKE
- Apache YuniKorn on Amazon EKS
Supported Autoscalers
- Amazon EMR Managed Autoscaling and Custom Autoscaling Policy on Amazon EMR
- Cluster Autoscaler and Karpenter on Amazon EKS
- Cluster Autoscaler with and without Node Auto-Provisioning (NAP) on Google GKE
Customer Optimization Stories
50% Cost Savings
“Pepperdata allowed us to significantly increase capacity for our Amazon EMR workloads and reduce our EC2 costs by over 50%. We can focus on our business, while they optimize for costs and performance.”
Improved Infrastructure Efficiency
“Pepperdata Capacity Optimizer helps optimize the efficiency of our infrastructure to reduce our spending.”
Automated Resource Optimization
“I was thinking there’s no way there’s a 30% overhead in [compute and memory resources] that everyone is paying that isn’t fixed.. But you see it in the numbers [with Pepperdata] that it’s true.”
Freedom From Manual Tuning
“[Pepperdata] freed me up to do all kinds of automation around cluster management, whereas [otherwise] I would have been day-to-day firefighting, figuring out where the resources were, what jobs were starving for resources, and then trying to play the resource police on the cluster.”
30-75% Cost Savings on Amazon EMR
“Saved 75% cost on [Amazon] EMR… Pepperdata Capacity Optimizer helped us to reduce the wastage on EMR clusters and increased the efficiency. For one of the clusters – the savings was as big as 75%. For others we are seeing at least 30% savings.”
Increased CPU Utilization
“[Pepperdata] Capacity Optimizer helps Pepperdata pay for itself… It’s allowing us to best utilize the compute we have available for our workloads.”
