The GCP Load Balancer is a software defined globally distributed load balancing service.
It enables GCP users to distribute applications across the world and scale compute up and down with very little configuration and cost.
Global Backend Service
Global – Backends(instance groups) that may be connected to this kind of backend service can be spread across multiple different regions.
Each backend is a resource to which a GCP load balancer distributes traffic. User should set 1 or more instance group per service.
Regional Backend Service
Regional – Backends(instance groups/target pools) that may be connected to this kind of backend service can be only from the region.
For regional backend service we utilize Scheme configuration to identify type of regional backend service:
An external backend service receives traffic from outside of the VPC (also called internet).
An INTERNAL backend service receives traffic from inside of the VPC (between the customer machines).
Backend Services With Elastigroup
If you are running an application or service behind GCP’s Backend Services (Load Balancers), Elastigroup provides the freedom to run blended cluster from multiple instance types, sizes
and purchasing options (
Elastigroup launches the target capacity for Preemptible VMs, to ensure a predictable and stable workload.
Its unique value proposition is its ability to add resiliency to workloads that otherwise would not be suitable to run as Preemptible VMs.
Integrating a backend service
- Enter the Creation Wizard to create a new Elastigroup or select Edit Configuration to integrate a backend service with an existing Elastigroup.
- In the Compute Tab under Backend Services click on ‘Add’.
- Set your required backend service (LB): Global / Regional.
- For Global just choose the backend service from the list.
You can add named ports, e.g.:
Ports: “[443, 8443]”
- For Regional choose the scheme type and than the backend service from the list.