Tag Archives: cluster

Enable VMware EVC by default on your new vSphere Clusters

If you don’t have a compelling reason to not enable EVC on a cluster, then I don’t see why you wouldn’t enable it. EVC allows you to future guard your cluster so that it allows you to introduce different CPU generations on a cluster.

There has been no evidence that EVC negatively impacts performance. But like with anything you should trust, but verify the application workload has the acceptable level of performance that is required.

There are impacts as to whether you have enabled EVC on an already existing vSphere cluster with running VMs. There is no outage if you do decide to turn on EVC on an existing cluster, but running VMs will not be able to take advantage of new CPU features until the VM is powered off and powered back on again. The opposite will occur if you decide to lower the EVC on an already EVC enabled cluster so the VM will be running at a higher EVC mode than the one that was explicitly lowered to until the VM is powered off and powered back on.

Enable_EVC

References:

vSphere Cluster Size

I’ve pondered over this question before, “how big should I make my cluster”. At that point you have to take a step back and think about the tradeoffs with the potential impacts that your next decision is going to bring.

So you have a design decision: Should you design a vertical or horizontal vSphere HA cluster?

At this point you have some design choices to make. Let’s assume that you have an already established cluster of 5 nodes with HA and DRS enabled. There is money in the budget to buy additional servers to accommodate future growth.

There is a considerable amount of information in order for you to consider in order to make the most appropriate decision, such as cost, power, cooling, floor space and max cluster limitations to name a few. There are advantages and disadvantages to designing a vertical or horizontal cluster.

Scale-up cluster

Advantage Disadvantage
Managing fewer host reduces administrative cost HA failover potential takes longer to complete
Less hardware to provide redundancy vice splitting the cluster in two and needing more host for failover Need to be careful to stay within the cluster VM pert Host maximums. Potentially resulting in VMs not being restarted after a failure.
More cluster resources reserved for failure
For DRS, fewer migration choices available to balance out the cluster
Patching large clusters can take longer

Scale-out cluster

Advantage Disadvantage
A host failure affects fewer VMs Potential affects the maximum size that a VM can be configured
HA failover takes less time More data center floor space
Fewer resources reserved for failover Increase costs, i.e. power, cooling
Less of a concern with staying within the cluster VM per Host maximums
For DRS, it provides a greater migration choice and more opportunities for a better workload balance

On top of these advantages/disadvantages then you have to decide if your current design is meeting the demand of the business to achieve the performance, scalability and the return on investments. Gathering a current state assessment will aid the decision making process to guide you towards a design. Getting back to the scenario above it definitely depends on a multitude of factors, but if you can identify all of these different interdependencies. There is not enough information to make a decision up front, but you can start to formalize a game plan to get you in the right direction.

With these choices being taken into account, then one must decide whether you’re meeting the requirement of the business needs. I definitely enjoy making those decisions and figuring out the best possible solution available.

Useful resources