Using predictive autoscaling | C2C Community

Using predictive autoscaling


Userlevel 7
Badge +65

When working with managed instance group (MIG), you can have an autoscaler automatically create or delete virtual machine (VM) instances based on increases or decreases in load. However, if your application takes several minutes to initialize, creating VMs in response to growing load might not increase your application's capacity quickly enough.

You can use predictive autoscaling to improve response times for applications with long initialization times and whose workloads vary predictably with daily or weekly cycles.

 

Limitations (until today)

  • Predictive autoscaling works only with CPU utilization as the scaling metric. Cloud Load Balancing or Cloud Monitoring metrics are not supported.
  • Compute Engine requires 3 days of CPU-based autoscaling history before it can generate predictions.
  • Predictions are based on weekly and daily load patterns. Compute Engine doesn't predict monthly, annual, or one-time events, and it doesn't predict load patterns that are shorter than 10 minutes. You can use schedule-based autoscaling to request capacity for one-time or other load patterns.

 

Has anyone used/checked/played with it?

@antoine.castex, @guillaume blaquiere have you tested it? What’s your opinion?

Google Cloud Blog

 


2 replies

Userlevel 6
Badge +15

For @antoine.castex and I, the cloud paradigm is the serverless. We don’t use (at least not so much and we avoid to use that) IaaS component and we can’t provide you more insight in these limitations and how to overcome them.

 

 

 

Userlevel 7
Badge +65

Thank you for your answer @guillaume blaquiere! I know you both love serverless solutions.

Is there anyone in our community who has tried it or intends to do so? 🤔

Reply