There are many technologies in the data center to keep workloads running even when hardware crumbles. One such...
technology is the fault tolerance feature of VMware vSphere that synchronizes duplicate VMs to ensure workloads continue to run unhindered.
Server hardware choices can either help or hinder VMware FT performance. Generally, it is good practice to either disable power management features on each host in the cluster -- such as power capping or forcing certain clock frequencies -- or be sure to set each server to the same power management configuration. If host servers allow features or portions of the instruction sets to be enabled or disabled in the firmware -- such as hyperthreading or Streaming SIMD Extensions 4, or SSE4 -- remember to configure each server's firmware the same way.
Also, consider the network implications of FT behavior. Logs are created and shared between original and duplicated VMs; this can inflate network traffic between servers within the cluster -- especially when servers host multiple original or duplicate VMs. For better performance, it may be necessary to upgrade LAN connectivity between the cluster nodes. For example, moving to multiple gigabit Ethernet ports or deploying 10 GbE connectivity between the cluster nodes -- often enabling jumbo frames -- can boost connectivity and alleviate latency between original and duplicated VMs within the cluster.
Ultimately, VMware FT offers a boost to traditional high-availability (HA) approaches. With HA, a failed VM will restart on another server -- leading to some disruption in workload availability. FT eliminates this disruption by running duplicate VMs kept in sync across a server cluster. But FT is not always easy to implement or configure. It's important to test any FT architecture to ensure that failover and recovery works without disruption. Clusters can also be sensitive to hardware, firmware and software differences, so replacing one physical server may have unforeseen consequences for cluster performance, leading to more server upgrades or replacements than otherwise expected.
Dig Deeper on VMware High Availability and Fault Tolerance
Related Q&A from Stephen J. Bigelow
Blade servers come in a variety of configurations. In order to effectively manage your data center, you'll want to consider your storage needs, blade... Continue Reading
When properly implemented -- and understood -- a cloud migration factory combines the right mix of people, processes and tools to smoothly transition... Continue Reading
Scaling up or scaling out is not a decision to be made lightly. Use monitoring data to determine the strategy that fits your use case and to inform ... Continue Reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.