Use Node Advisor recommendations to save money
Most of the time you can save on your cloud bill. Magalix does the analysis of your VMs and recommends different ways to save on your cluster. Savings can be done by adjusting your overall capacity, changing VMs types/count, or selecting a different billing model. Magalix checks different options for your cluster considering current and expected workloads.
Step 1 - Understand Cost-Saving Recommendations
Cost-saving recommendations focus on reducing your overall cloud bill based on the following factors:
- The current and expected usage of CPU and memory. It uses the collected history and utilized prediction models to estimate the needed capacity for the next few hours or days.
- The currently used VMs types and sizes. Magalix checks if the used VMs match the workloads or not to suggest better overall capacity.
- Cluster sizing policies, which you can set yourself based on your business and technical needs.
A recommendation will tell you the required action and all relevant context around it.
The required action will mostly provide instructions to change nodes types, sizes, and/or numbers. It depends on the above-listed factors, the suggestion will provide a detailed plan of the new cluster size that you need to apply.
This section tells how many yearly savings you will achieve if you apply those recommendations. This saving is based on on-demand pricing. It also uses the current cluster size as the baseline for that calculation. You will see more detailed analysis in the capacity table to compare other billing plans.
This section explains what the Magalix agent was able to observe about the current capacity and compares it to the current cluster capacity. It provides a detailed analysis of the current and suggested utilization, capacity, VM types, and different billing plans. Please read our Node Advisor detailed spec for more details about the supported cloud providers, billing plans, etc.
This section shows the aggregated usage, allocated, and currently available of CPU and memory. It provides a visual comparison to help you decide if you are under or over-utilizing the capacity that you allocated for your cluster.
How to Resolve the Issue
Magalix Node Advisor does not currently provide automation of VMs resizing and allocation. You will find in this section different ways to apply those changes yourself.
Please let us know if you need this to be automated
Changing the capacity allocation at the infrastructure level takes a lot of hand-holding and integration with the cloud provider. Please drop us a line to understand your needs and prioritize them accordingly.
Resources to Learn More
If you would like to learn more about resource allocation inside Kubernetes and how to manage them, you will find this section handy. We are linking to official Kubernetes documentation and also our more detailed explainer articles.
Step 2 - Apply Recommendations and Save Money
Magalix does not provide automation yet to apply VM level changes. You will need to apply those changes yourself based on your needs.