Auto Scaling

Auto Scaling

Autoscaling is a cloud computing technique for scaling cloud-based services and dynamically allocating computational resources. Depending upon the user needs, server traffic, and load on a server, the active number of servers (typically called workers) varies automatically. Typically, several workers are activated simultaneously as required when the load peaks, while when there is no substantial traffic on a server, the workers are scaled down (switched off). Hence, the only optimum number of workers are active at any time.
In Rubiscape, the autoscaling settings are tenant-specific. As you change the tenant, these settings also change. It may so happen that one of your projects with a specific workload is handled through one tenant. Hence, separate settings may be required and can be configured for separate tenants.
Currently, two scenarios are possible related to auto-scaling in Rubiscape.

Auto-scaling disabled:

  • NO auto-scaling
  • Worker node configurations are default and are decided by the admin

Auto-scaling enabled:

Rubiscape provides support for the following two cloud providers:

  • Microsoft Azure
  • AWS

The following four drop-downs are the same for Microsoft Azure and AWS.

  • Maximum Workers (Frequency)
  • Number of tasks per worker
  • Instance Type
  • Instance