GCP network passthrough Load Balancer Multiple Front Ends

[needlestack]

What is the behavior of GCP load balancer with respect to :
- network pass through load balancer
- multiple frontend rules configured

Since health check is sent to each frontend ip - how the loadbalancer determines when a backend is healthy or no. Like if the backend responds to probe on frontend ip 1 but not frontend ip2.

[princesslyca]

Hello @needlestack

Welcome to Google Cloud Community Support!

Passthrough Network Load Balancers are like traffic directors for your servers. They distribute incoming traffic to different servers within the same region without altering the original information in the data packets. This means the servers can interact directly with the clients, allowing for a wider range of protocols and giving you more control over SSL certificates.

Health checks are based on a series of probes, and the backend's overall health is determined by the number of successful or failed probes. Multiple independent probers are used to ensure continuous health monitoring, even if one prober fails. A backend is considered healthy if it responds successfully for the configured number of times, and unhealthy if it fails to respond for a predetermined number of attempts. See Health Check Overview

In addition, to estimate the probe frequency per backend service, consider the following:

• Multiple forwarding rules for internal passthrough Network Load Balancers. If you have configured multiple internal forwarding rules (each having a different IP address) pointing to the same regional internal backend service, Google Cloud uses multiple probers to check each IP address. The probe frequency per backend service is multiplied by the number of configured forwarding rules.
• Multiple forwarding rules for external passthrough Network Load Balancers. If you have configured multiple forwarding rules that point to the same backend service or target pool, Google Cloud uses multiple probers to check each IP address. The probe frequency per backend VM, is multiplied by the number of configured forwarding rules.

Google Cloud uses the following configuration flags to determine the overall health state of each backend to which traffic is load balanced.

See Destination for probe packets.

Hope this helps.

[needlestack]

Hi and thanks for your reply but that did not answer the question. Please see the following comments:

One) You have multiple frontends each one with a different ip. A probe will be sent to each frontend ip and that frontend ip will be configured on the backend vm.  If the vm stops responding on frontend ip1 will the loadbalancer Mark as unhealthy for frontend ip2.  

I don't think it will do that because it is a big problem.

Two) in gcp load balancing there is no concept of a load balancing rule like Azure. You configure a frontend and a backend that is right but by default they are all associated with each other. Go configure a load balancer from console what you do it :

Configure multiple frontends 

Configure backends after that step

But there is no step where you associate a frontend with a specific backend.

Thanks.