NSX LB: Always favoring one server over another

[vmmedmed]

Say you had two servers in a pool Server A and Server B reachable at port 8080. In all instances you want

requests to go to Server A unless Server A stops responding to keepalives on port 8080. If I set

Server A to weight 200 and Server B to weight 1 and use algorithm URI - would that do it? If not

what's the mojo?

[bayupw]

If you are looking for an Active/Standby members, then you would need to use application rule and create 2 different pool.

As far as I know, there is no out of the box features for Active/Standby member or Active/Standby pool

The steps would be

1. Create 2 pools, pool_serverA with member serverA and pool_serverB with member serverB

2. Create Application Rule as below

acl poolServerA_isDown nbsrv(poolServerA) eq 0

use_backend poolServerB if poolServerA_isDown

3. Apply the Application Rule to the Virtual Server

If you see this blog post: Configuring NSX-v Load Balancer for use with vSphere Platform Services Controller (PSC) 6.0 - VMware...

It is using similar Active/Standby Application Rule

The first application rule will be used by the HTTPS virtual server to switch between the corresponding pools for the HTTPS backend servers pool.

# Detect if pool "pool_psc-01a-http" is still UP 
acl pool_psc-01a-http_down nbsrv(pool_psc-01a-http) eq 0
# Use pool " pool_psc-02a-http " if "pool_psc-01a-http" is dead
use_backend pool_psc-02a-http if pool_psc-01a-http_down

[bayupw]

Just posted a blog post for your reference here: living in the cloud cadence: Creating an Active-Passive Pool on VMware NSX Load Balancer

[vmmedmed]

This worked like a champ. Thank you very much.

[bayupw]

Glad it worked for you.

If you find my answer useful please consider to mark it as Helpful or Correct