I have two esx hosts in a cluster
There are 20 VM's running utilizing 2 ghz and 5 GB of memory total.
Each esx have 16GB ram and 20 GHZ capacty
I still get current failover capacity 0 hosts.
My VM's are organized in resourcegroups, and the two top level groups have resource settings:
test:
-
CPUreservation 3500 Mhz,limit 3500Mhz,
-
Memory reservation:2000MB, limit 3000MB
operations:
-
CPUreservation, 13000Mhz, Limit 16000Mhz
-
Memory reservation, 12.000MB, limit 16.000MB
My VM's are all sett to reservation 0, limit: unlimited
So i think the reason for this is that the failover capasity calculations is based on the ressource pool limit.
So the total memory is 19 gb and you only have 16 GB pr. host.
Try to set the Productions memory limit to ex. 12gb
/Lars
I do not think it is becaue the limit is too high - but you might check to see if that is it by changing the RP limits to unlimited - if that is it you should get a failover capacity of 1 but I do not think that is it -
The way failover cpapacity is calculaated is by determining the number of slots available - this is calculated by using vm reservations - see for details
BUT you do not have reservations set on your VMs so how could this apply - I think what might be going on in older versions of ESX if no reservation was set the vmekernel assumed a reservation of 1/2 of the available memory to see if the vm could start - so I am thinking in calculating failover capcity vmware uses that number when the vm has no reservation set so - looking at your configurations your just at the limit for CPU and over for memory - if this were true - to test set a nominal resrvation - 100 MHz for CPU and 126 MB for memory - see if that changes -