From a2a65fd7298cd4764d4928b80a609df3e565835e Mon Sep 17 00:00:00 2001 From: Angela Costa Date: Tue, 17 Dec 2024 10:06:07 +0000 Subject: [PATCH] Adds info about failback behavior --- .../steering-policies/standard-options.mdx | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/src/content/docs/load-balancing/understand-basics/traffic-steering/steering-policies/standard-options.mdx b/src/content/docs/load-balancing/understand-basics/traffic-steering/steering-policies/standard-options.mdx index 5ef21f88320171..4d3a8564a1ba54 100644 --- a/src/content/docs/load-balancing/understand-basics/traffic-steering/steering-policies/standard-options.mdx +++ b/src/content/docs/load-balancing/understand-basics/traffic-steering/steering-policies/standard-options.mdx @@ -23,6 +23,16 @@ If all pools are marked unhealthy, Load Balancing will direct traffic to the fal If no monitors are attached to the load balancer, it will direct traffic to the primary pool exclusively. +### Failback behavior + +In an active/standby setup, with two origin pools: + +- Traffic always routes to Pool 1 (the primary pool) unless it becomes unhealthy. +- If Pool 1 is marked unhealthy, traffic shifts to Pool 2 (the standby pool). +- Once Pool 1 becomes healthy again, traffic automatically shifts back to Pool 1, assuming no [session affinity](/load-balancing/understand-basics/session-affinity/) or other settings require subsequent requests to stay at Pool 2. + +This behavior is known as failback and ensures traffic resumes normal routing when the primary pool recovers. + ## Random steering Choose **Random** to route traffic to a healthy pool at random. Customers can use this option to set up [active - active failover](/load-balancing/load-balancers/common-configurations/#active---active-failover) (also known as round robin), where traffic is split equally between multiple pools.