Feature description
All urltest / auto-select groups share a single global health-check URL (the app's URL-test setting), and the pool is the whole subscription. There is no way to give each auto-select group its own health-check target or its own subset of nodes.
Please allow, per auto-select group:
- a custom health-check URL (the target that defines "healthy" for that group), and
- a custom node pool / subset (which nodes participate in that group).
Use case
Different services need different reachability targets and different node pools. One service should auto-select the lowest-latency node that can reach service A's API, another against service B's API; and a geo-restricted service should auto-select only from a curated set of allowed-region nodes.
Today, with one global URL-test target and the whole subscription as the pool, a single "Auto" group can't represent these per-service needs — users must pin one node manually (no failover) or build external tooling.
Combined with per-group expected-status (see #1959), per-group health-check URL + node subset would enable true per-service automatic failover — keeping each service reachable by auto-picking a node where it actually works.
Target OS
All platforms.
Feature description
All
urltest/ auto-select groups share a single global health-check URL (the app's URL-test setting), and the pool is the whole subscription. There is no way to give each auto-select group its own health-check target or its own subset of nodes.Please allow, per auto-select group:
Use case
Different services need different reachability targets and different node pools. One service should auto-select the lowest-latency node that can reach service A's API, another against service B's API; and a geo-restricted service should auto-select only from a curated set of allowed-region nodes.
Today, with one global URL-test target and the whole subscription as the pool, a single "Auto" group can't represent these per-service needs — users must pin one node manually (no failover) or build external tooling.
Combined with per-group
expected-status(see #1959), per-group health-check URL + node subset would enable true per-service automatic failover — keeping each service reachable by auto-picking a node where it actually works.Target OS
All platforms.