Zevenet 5.13 - HTTP FARM - Farmguardian 'disabled' not work

[dboc]

Hello,

We have verified that Zevenet 5.13 version always check backend health even when Farmguardian 'disabled' is used.

This behavior is wrong, I have a use case that I don't need zevenet to check the backend, because the service is only active from time to time. When I trigger this service, zevenet waits a long time to verify the integrity and forward the request to the correct backend.

I just want that Zevenet do not verify, always forward the request even if the backend is unhealthy.

To replicate this, just create an HTTP FARM, a "disabled" Farmguardian, add a backend. On the backend server, UP the service and try to access it instantly, zevenet will reject all request until it checks the integrity of backend, it takes about 5 seconds.

[nevola]

Hi @dboc, this is the default behavior for the proxy (http farm). You can use either the Layer 4 or decrease the seconds to check the resurrected backends in the HTTP Farm.

Cheers.

[dboc]

@nevola
I understood, I thought it was an error, but you said that is built in designed.

For me Layer 4 it is not possible, we have to user proxy(http farm) because it looks the request host and path to redirect to differents backends.

Adjust resurrected timeout needs to adjust backend connection timeout also. And we need to ressurected timeout be close to zero, and backend connection timeout could not be close to 0.

It is possible to make an option in proxy(http farm) to not verify backend health at all?

Ty