Predicting the growth of resources in your Kubernetes deployments

This use case describes how to use BMC Helix Continuous Optimization to predict the growth of your resources and how to avoid the risk of saturation.

Scenario

Teresa is a Kubernetes Technology Specialist and plans to onboard a new tenant on July 1. She wants to analyze the capacity of the existing resources and identify the potential risks and bottlenecks in the existing Kubernetes deployments.

Teresa wants to simulate the growth of resources and see the impact of the change on the Kubernetes resources. She faces some of the following challenges in her IT operations management:

  • Forecasting the growth of resources based on historical data.
  • Identifying the resources that will be saturated over time.

Teresa needs an effective solution to identify all the resource capacity and to quickly identify the resources that are nearing saturation, and to take action to avoid the risk of saturation.

Implementation workflow

Teresa requests that the tenant administrator Tim complete the following activities so that she can predict the capacity of the resources by using BMC Helix Continuous Optimization.

Task 1: To view the Kubernetes deployments

Teresa needs to do the following steps to view the health of the Kubernetes resources:

  1. Log in to BMC Helix Continuous Optimization.
  2. From the Views tab, navigate to Containers > Kubernetes.
  3. Click the Controllers tab and then click the Capacity tab to view the resource utilization of the controllers as shown in the following example scenario:


Insights for Teresa

  • The CPU utilization is low compared with the limit and request.
  • The memory usage is moderate and it currently does not indicate a risk of saturation.

For more information about the details displayed in the Kubernetes view, see Kubernetes view.

Task 2: To predict the growth of resources and avoid the risk of saturation

Teresa needs to do the following steps to analyze the health of the resources in the future when she plans to onboard a new tenant:

  1. Select the controllers for which you want to simulate the growth.
  2. Click What-if simulation.
  3. In the Increase percentage field, add a 30% increase on July 1.
  4. Click Simulate.


Insights for Teresa

  • From the results, it is evident that one or more resources will be saturated before the expected growth on July 1.
  • The timeline indicates that with the current resources, there is no risk of saturation until the end of June.
  • From the Deployment list, it is clear that the CPU and Memory usage of one deployment is nearing saturation and requires immediate attention because the levels can impact the usage of these resources on July 1. 
  • Other details, such as bottleneck, CPU and memory utilization at the end of the forecast period help to identify the real cause of saturation and resolve the issue.
  • The Days to saturate value indicates how many days before the resource will be saturated. This indicator helps in adding resources on that date to avoid saturation.

For more information, see Predicting the growth of your resources by using simulations.

Task 3: To increase the resources and add changes to the forecast

To address the saturation issue, before the end of June, Teresa will need to add resources.

  1. Click Add change.
  2. Select the deployment that shows a risk of saturation.
  3. Click Scale out and add CPU cores and memory on July 1 to the deployment.

Task 4: Check the results and share the simulation

Based on the changes added, the simulation is rerun and the forecast changes. At a glance, the timeline shows green for the entire forecast period. After Teresa sees that the results indicate that there is now no risk of saturation, she must complete the following steps:

  1. Check the status of the listed deployments and confirm the status of the resources.
  2. Save the simulation and send the link to the Infrastructure team to view the additional resources that should be added.

Insights for Teresa

  • Review the changes in the Resources required section.
  • There is no more a risk of saturation. The CPU and memory utilization are normalized.

Results

Teresa is able to solve the saturation issue and she can onboard the new tenant in the future based on the simulation results. 

Benefits

Teresa can find answers to most of her challenges by using BMC Helix Continuous Optimization. Using the What-if simulation feature, Teresa can:

  • Identify the resources at risk of saturation.
  • Determine the most probable causes of resource saturation.
  • Based on past behavior, predict the growth of resources.
  • Plan to add resources in a particular deployment before the date of saturation to avoid the risk of saturation.
Was this page helpful? Yes No Submitting... Thank you

Comments