The (public facing) Azure Load Balancer that I've setup through the Azure portal using default values points to a pool of 3 Virtual Machines.
On each of these VMs, a tiny node.js server is running, only responding a 200 HTTP status code to the / path each time the default health probe call it every 5 seconds. A log on stdout confirm that each calls is handled correctly, and it is confirmed on the load balancer side which detects all instances as healthy.
Now, using the health probe form from the portal, I change the interval value of the load balancer for probe from 5 to 30 seconds and save this configuration. I've refreshed the portal page to make sure that the 30s value was indeed saved.
Unfortunately, it appears that the load balancer totally ignores this new value and continue to probe the VMs every 5s instead of every 30s I expected. This behavior is confirmed by the logs I have on each VMs.
What am I missing here? Is there an extra step needed to ensure that the new interval value is effective?
