DanieleFiore201
Enthusiast
Enthusiast

HA current failover capacity 0 host

Jump to solution

Hi guys,

in this cluster i'm hosting veru big VM with reservation on VM level.

I have 4 x host with 256 GB RAM each.

HA cluster settings "cluster tolerates 1 host failure" and i have few slot with the following  :

4 slot in the cluster

4 used slot

0 available slot

0 failover slots

host size :

256mhz

10vCPU

132119MB

i'm a little bit confused about such "zeroes" especially the failover slot......

Does it mean cluster can still tolerate 1 host failure or not ?

Time ago i got an good explanation from the communities , see below.....

But my question is also is used slot - available slot =>always the sum of the value indicated in failover slot ?

Example:-

Total Number of slots = 22

Used Slots = 4 (Depends on Numbers of VM Running)

Available Slots = 7

Failover Slots = 11 ( Depends on How many hosts failure cluster can tolerate) if it is 1 and we have 2 node cluster then it is 50% of the total slots. Means you can use it according to this example you can only power on 11 VM's now not more than that.

Many thanks in advance.

Daniele

0 Kudos
1 Solution

Accepted Solutions
depping
Leadership
Leadership

Daniele Fiore wrote:

H Ranjna

i got it right but can you answer to my question below ? actually that was an asnwer from you on a previous post :smileygrin:

Can you explain these values :

Total slots in the cluster : 8

Used slots : 4

Available slots : 2

Failover slots 2

Admission control policy : 1 host crashes

So as things are now what does it mean :

That HA will start only 2 VMs in total ? or 2 +2 (available slots + failover slots)

And related to "f the failover slots are zeros then cluster cannot tolerate any host failure" does it mean =>HA will restart NO VMs at all or simply HA will restart only some of them?

Many thanks in advance,

Daniele

You have 8 slots in total in your cluster

4 of those slots are in use by virtual machines

You can still power on 2 more virtual machines

And 2 slots are reserved in total as "capacity" to allow for a fail-over.

In this case lets assume you have 4 hosts. Each hosts can cater 2 slots. When you set admission control policy to 1, it will subtract 2 slots (amount of s single host) from the total amount of slots in the cluster. (8-2)

View solution in original post

0 Kudos
9 Replies
RanjnaAggarwal
VMware Employee
VMware Employee

if the failover slots are zeros then cluster cannot tolerate any host failure

Regards, Ranjna Aggarwal
0 Kudos
DanieleFiore201
Enthusiast
Enthusiast

H Ranjna

i got it right but can you answer to my question below ? actually that was an asnwer from you on a previous post :smileygrin:

Can you explain these values :

Total slots in the cluster : 8

Used slots : 4

Available slots : 2

Failover slots 2

Admission control policy : 1 host crashes

So as things are now what does it mean :

That HA will start only 2 VMs in total ? or 2 +2 (available slots + failover slots)

And related to "f the failover slots are zeros then cluster cannot tolerate any host failure" does it mean =>HA will restart NO VMs at all or simply HA will restart only some of them?

Many thanks in advance,

Daniele

0 Kudos
rickardnobel
Champion
Champion

Daniele Fiore wrote:

HA cluster settings "cluster tolerates 1 host failure" and i have few slot with the following  :

4 slot in the cluster

4 used slot

0 available slot

0 failover slots

host size :

256mhz

10vCPU

132119MB

i'm a little bit confused about such "zeroes" especially the failover slot......

This means that from HA point of view you have space for total 4 VMs in your cluster and you have these 4 VMs already powered on, which leaves no spare capacity. The 0 failover slots means that HA can not guarantee that all VMs will be able to power on, due to the very large reservations - which has caused so large slot sizes.

My VMware blog: www.rickardnobel.se
RanjnaAggarwal
VMware Employee
VMware Employee

Can you attach the snapshot of this configuration or this is just an imaginary figure

Regards, Ranjna Aggarwal
0 Kudos
depping
Leadership
Leadership

Daniele Fiore wrote:

H Ranjna

i got it right but can you answer to my question below ? actually that was an asnwer from you on a previous post :smileygrin:

Can you explain these values :

Total slots in the cluster : 8

Used slots : 4

Available slots : 2

Failover slots 2

Admission control policy : 1 host crashes

So as things are now what does it mean :

That HA will start only 2 VMs in total ? or 2 +2 (available slots + failover slots)

And related to "f the failover slots are zeros then cluster cannot tolerate any host failure" does it mean =>HA will restart NO VMs at all or simply HA will restart only some of them?

Many thanks in advance,

Daniele

You have 8 slots in total in your cluster

4 of those slots are in use by virtual machines

You can still power on 2 more virtual machines

And 2 slots are reserved in total as "capacity" to allow for a fail-over.

In this case lets assume you have 4 hosts. Each hosts can cater 2 slots. When you set admission control policy to 1, it will subtract 2 slots (amount of s single host) from the total amount of slots in the cluster. (8-2)

View solution in original post

0 Kudos
rickardnobel
Champion
Champion

Daniele Fiore wrote:

Can you explain these values :

Total slots in the cluster : 8

Used slots : 4

Available slots : 2

Failover slots 2

Admission control policy : 1 host crashes

This is likely a cluster with 4 hosts and the "slot size" is so large that it only allows room for two VMs on each host. With the default Admission Control Policy = host failure cluster tolerates 1, this will make HA try to stop you from powering on too many VMs while everything is working.

Since you have 4 hosts with 2 "slots" and should be able to tolerate that one of these hosts go down then it will allow you to power on two more ("available slots"), while two slots are hold back to be used in a crash ("failover slots").

My VMware blog: www.rickardnobel.se
ancoleman
Contributor
Contributor

I need a little incite from some experts. I do not have admission control enabled. However as you can see from the first two clusters everything appears to be in order. However, in our third cluster (production) our number is way off and the Slot size is way different in terms of memory. Here is the details of that cluster:  11 Hosts, 4 hosts have 144GB memory 24 logical procs each, 6 hosts have 96GB memory 24 logical procs each, 1 host 256GB memory 40 logical procs. Now this cluster is bigger in terms of memory and cpu versus our other clusters, but the Slot size differs dramatically. So my question is why? How can I get this sorted out? Thanks in advance.

Scenario:

Cluster 1 Non Prod:

Screen Shot 2013-01-31 at 8.57.55 AM.png

Cluster 2 Prod Pci:

Screen Shot 2013-01-31 at 8.58.11 AM.png

Cluster 3 Production:

Screen Shot 2013-01-31 at 8.57.02 AM.png

0 Kudos
ancoleman
Contributor
Contributor

I think I found the answer to my question here:

http://communities.vmware.com/thread/270579

0 Kudos
lenzker
Enthusiast
Enthusiast

you have a huge slot-size @ RAM: What is the highest RAM reservation that you have set?

Go to the resource allocation tab in your cluster and check what the highest reservation!

VCP,VCAP-DCA,VCI -> https://twitter.com/lenzker -> http://vxpertise.net
0 Kudos