Managed Instance Group
A group of identical Compute Engine VMs created from an Instance Template and managed as a single unit, GCP's equivalent of an AWS Auto Scaling Group or Azure VM Scale Set. A Managed Instance Group can autoscale based on a metric like CPU utilization, adding or removing instances automatically within configured minimum and maximum bounds.
Frequently Asked Questions
What's the difference between a zonal and a regional Managed Instance Group?
A zonal MIG places all its VM instances in a single zone, which is simpler but vulnerable to a full zone outage. A regional MIG spreads instances across multiple zones within a region automatically, and GCP recommends regional MIGs for production workloads because they survive a single zone failure without manual intervention. Both are driven by the same Instance Template, so switching doesn't change how instances are configured, only how they're distributed.
What's a common mistake when configuring autoscaling on a Managed Instance Group?
Relying solely on CPU utilization as the scaling signal is a frequent misstep — a workload that's I/O-bound or memory-bound can be under real load while CPU stays low, so the group never scales out in time. It's usually better to combine CPU with a custom Cloud Monitoring metric (queue depth, request latency) and to set a sane cooldown period, otherwise the group can thrash — scaling up and down repeatedly — during bursty traffic.