shredles
Contributor
Contributor

HA Failover Capacity in ESX 3.5

Can someone explain how HA failover capacity calculations are done in ESX 3.5? I am seeing a situation where I clearly have plenty of available memory and CPU to handle a single server failure based on the reservations I have set, but HA reports my failover capacity at 0.

Specifically, I have 3 hosts, 25 guests, and my cluster resource allocation values look like this:

CPU Reservations: 69564MHz

CPU Reservation Used: 6388 Mhz

CPU Unreserved: 63176 Mhz

Memory Reservation: 89692 MB

Memory Reservation Used: 16550MB

Memory Unreserved: 73142 MB

Oddly, If I change my cluster to use strict admission control, my cluster turns red and my cluster resource allocation values change to this:

CPU Reservations: 69564MHz

CPU Reservation Used: 24388 Mhz

CPU Unreserved: 45176 Mhz

Memory Reservation: 89692 MB

Memory Reservation Used: 27647MB

Memory Unreserved: 62045 MB

Why do these values change when I change admission control if my reservations haven’t changed? Even with these values listed my cluster should be able to handle a server failover with resources to spare.

0 Kudos
0 Replies