Skip to Content
Core InfraWorkload RightsizingWorkloadsWorkload Overview

Workload Overview

To see supported workloads, please refer to the Workload page.

In order to get full visibility of a workload, click on the workload under the Overview page.

The Workload Overview provides a comprehensive view with the following tabs:

Pod Rightsizing

Resources

When ephemeral storage optimization is enabled, ephemeral storage metrics appear alongside CPU and memory.

  • usage (avg) - Average resource usage across workload replicas
  • usage (p90) - Percentile 90 of resource usage across workload replicas
  • usage (max) - Max of resource usage across workload replicas
  • optimized request - A simulation of the recommended request at any given time while taking into account the selected Policy
  • current request - The resource request of the pods at any given time
  • original request - The resource request before the ScaleOps optimization
  • current limit - The resource limit of the pods at any given time

You can select different policies by clicking on the Policy name to see what-if scenarios and get understanding of the recommendation that will be provided for policies.

Timeline

Under the Timeline you can troubleshoot and get visibility on the events that are related to workload.

Replicas Available in v1.12.0+

Under the Replicas tab you can get full visibility on the Replicas Optimization for the workload.

Learn more about Replica Optimization here.

Troubleshoot

Under the Troubleshoot tab you can get full workload visibility to troubleshoot the workload. Leverage one of the default dashboards or create your own by selecting the desired charts.

Charts

  • CPU, Memory, Ephemeral storage (when enabled) - Resources Usage, Request, Recommendation and Limits
  • Automated - When was the workload automated
  • Replicas - Track the number of ready replicas compared with the workload’s desired replicas
  • Node CPU utilization - Track the CPU utilization of nodes running the workload’s pods. Nodes with CPU high-utilization can cause throttling
  • Node Memory utilization - Track the Memory utilization of nodes running the workload’s pods. Nodes with Memory high-utilization can cause Out-of-Memory events
  • CPU Throttling - Track CPU throttling events across workload’s replicas
  • Out of memory - Track all Out-of-Memory events
  • Liveness probe failures - Track workload’s livenessProbe failure events
  • ScaleOps update evictions - When ScaleOps has initiated an eviction
  • HPA CPU - Track the HPA CPU utilization
  • HPA Memory - Track the HPA Memory utilization
  • CPU noisy neighbors - under-provisioned workloads - Track amount of workloads that were cpu-stressed while other pods on their node used CPU above request.
  • Memory noisy neighbors - under-provisioned workloads - Track amount of workloads that were memory-stressed while other pods on their node used Memory above request.
  • Pod CPU request - Track the CPU request of the pods related to the workload
  • Pod Memory request - Track the Memory request of the pods related to the workload
  • Nodes life cycle - Track the amount of spot/on-demand nodes this workload was running on.

Pods Available in v1.15.15+

Under the Pods tab you can get full visibility on the pods that are part of the workload.

GPU (Coming soon)

APIs Available in v1.11.20+

Under the APIs tab you can get full visibility on the APIs that are part of the workload.

Learn more about API Observability here.

Network Available in v1.5.37+

Under the Network tab, you can gain insights into the network traffic associated with this workload, including communication patterns, traffic volumes, and costs.

Learn more about the Workload Network view here.

Events

Under the Events tab you can get full visibility for decisions and events that were taken place by the ScaleOps platform for this workload. All events are k8s native events events.k8s.io/v1.

YAML

Under the YAML tab you can get full visibility on the YAML of the workload and if workload is managed by an HPA, you can see the YAML of the attached HPA Available in v1.11.20+.

What’s Next?

Policies

WorkloadsWorkload Actions