The impact that virtualization has made on IT has been profound and continues to grow year over year. Fueled in part by the substantial growth of cloud-based services, many industries have been eager to migrate their existing server infrastructure to high-density, small-footprint hardware to consolidate their physical server footprint, while lowering costs specific to managing and maintaining larger server farms.
SEE: WI-FI 6 (802.11AX): A cheat sheet (free PDF) (TechRepublic)
All hypervisor vendors offer tools to easily convert physical servers to virtualized images that can then be imported to the host and run with a few clicks of the mouse. The process to run these virtual instances does not end there: While simply performing these basic steps will yield a working virtual machine (VM), it takes more hands-on effort to configure optimized VM settings.
By keeping an eye on performance and tweaking the settings necessary to take advantage of the host server’s resources, VMs can be configured to perform capably and efficiently. Furthermore, by modifying resource utilization per instance, the virtual host will work better as well, perhaps even freeing up enough resources to host several more instances than were previously in use due to lack of optimization. Read the following tips on how to boost VM performance and maximize efficiency.
1. Perform a baseline assessment
One of the biggest payoffs for IT in saving time and money is to assess the use of resources, also known as performing a baseline. This assessment should be performed over a period of time, for example over two weeks, and serves to provide insight into what resources the server is actually using, compared with what the physical server is configured for.
The baseline should also be taken several times a day during this timeframe, specifically during peak and off-peak times so IT can best configure the resources on the VM so that it’s not under-provisioned (the VM performs poorly) or over-provisioned (too many resources are sitting idle).
There are apps that will monitor servers for resource utilization and provide detailed reporting, though operating systems have built-in performance analyzers that provide this information as well.
2. Remove unnecessary software
Servers running on bare-metal hardware require specialized software or drivers like all computers. However, their virtual counterparts do not require such software to operate normally. By keeping device drivers or vendor-specific or proprietary software on these VMs, it can lead to abnormal behavior and instability.
To prevent this from affecting your VMs, it’s a best practice to uninstall all device drivers that are no longer being used, along with any software that is not required by the hypervisor’s vendor or to run the services on the VM. Depending on your department’s preventive maintenance policy, you may be surprised at just how many resources will be freed up on the VM and, subsequently, the host server.
3. Rightsize configuration settings
The two steps above will help IT make the best-informed decisions when it comes to this third step. Rightsizing a VM involves configuring the settings of each instance so that it uses what’s necessary with a bit of padding to account for spikes in performance. This helps keep both the VMs and host server resources in check so neither gets affected in the event of a resource spike.
By helping identify exactly what resources are being used, IT can determine how many VMs will run on the host properly and efficiently. These processes will vary from instance to instance and depend on what the VMs will be used for. But the initial configuration, set up effectively, will allow for the instances to function with little trouble.
Consider this process cyclical because over time, a VM’s purpose might shift and change based on the organization’s needs. Continuing to monitor baselines, adjusting for changes in software requirements, and reconfiguring host settings per VM will keep your virtual environment operating smoothly.