Google Cloud's Managed Instance Groups (MIGs) automatically scale your compute capacity based on demand, ensuring your application maintains performance during traffic spikes while minimizing costs during quiet periods.
Configuring Scaling Policies
Autoscaling policies can trigger on CPU utilization, HTTP load balancing metrics, Stackdriver custom metrics, or Pub/Sub queue depth. Combining multiple signals with appropriate cooldown periods prevents oscillation between scaling events.
Instance templates define the machine type, boot image, startup scripts, and metadata for every instance in the group. Updating a template and performing a rolling update ensures zero-downtime deployments as old instances are replaced with new ones.
Regional MIGs distribute instances across multiple zones within a region, providing automatic redundancy against zone-level failures. Combined with a global HTTP(S) load balancer, your application achieves both geographic reach and fault tolerance.