Get started Bring yourself up to speed with our introductory content.

Improving performance: Virtualization management guide

In virtual environments, managing performance now requires more sophisticated techniques. This guide offers resources and advice on topics including setting benchmarks, understanding server performance requirements for a particular environment, VirtualCenter's mapping feature to depict your infrastructure and more.

In virtual environments, managing performance with old standby server metrics like CPU and memory usage won't cut it anymore. These increasingly dynamic and fluid environments don't lend themselves to isolating root causes. Nevertheless, infrastructure managers can find an alternative to simply measuring isolated metrics and achieve the true goals of performance management: ensuring uptime, improving application performance and maximizing resources in their data centers.

This guide introduces virtual infrastructure managers to the new requirements of performance management and offers advice on which practices, products and metrics truly matter in assessing one's environment. Below we feature advice on minimizing virtual sprawl, setting performance benchmarks, analyzing system metrics, mapping your virtual environment, minimizing downtime and maximizing resources.


Virtualization systems management: Performance

  Virtual machines

  Virtual servers

  Virtual storage

  Analyzing statistics

  Virtualization platforms

Virtual machines

Benchmarks and metrics for measuring virtual machine performance
Performance measurement and reporting is critical to the success of a virtualization implementation. Tracking virtual machine (VM) performance to pinpoint problems, or just to have meaningful reports of resource consumption, can be complicated because a virtual machine's behavior is strictly related to the underlying host. This is further complicated because performance depends heavily on what other virtual machines are doing. This tip breaks down virtualization performance measurement into the essentials.

The effects of virtual machine sprawl on performance
Virtual sprawl -- the proliferation of virtual servers -- has become an increasing a concern because it's so easy to create and deploy new servers. As a result, organizations seeking to reduce complexity and costs by consolidating their infrastructure through virtualization may ultimately confront a different version of the same basic management issues they attempted to resolve in the physical universe.

Best practices for preventing virtual machine sprawl prevention
Preventing VM sprawl will contribute to better overall performance of virtual machines. The ease of deployment can make headaches out of too many virtual machines on a host. By treating virtual machines like physical machines, ensuring that all VMs are monitored and updated regularly and by deploying only base images and supported configurations, administrators can effect an even more efficient virtual environment.

Virtual servers

Performance optimization on virtual servers involves networking and availability
This tip explores some of the issues frequently presented in managing network resources and automation strategies in virtualized server environments. Assigning priority to servers for a high-availability virtual environment is one way to leverage a virtual server for optimal performance. System management of virtual server performance also requires network redundancy and isolation. For a virtual implementation, planning is the most important step. For networking and automation configuration, the key is to define the expectation for the technology and assign the parameters in which the virtual environment must perform.

Planning prevents slow virtual servers
Virtualization expert Andrew Kutz explains how to fix a slow response on a virtualized Web server running on VMware ESX. According to Kutz, the problem has nothing to do with ESX but with the inevitable gotchas of a first-time virtualization project and the need to test the correct server configuration for one's environment.

Several methods for troubleshooting performance in virtual servers
Virtualization-related overhead can't be boiled down to a single number or percentage. Rather, it varies significantly with the specific resource usage needs of a VM and the applications it's supported. Administrators should focus on the end-user experience and measure performance. This often involves simulating transactions or common operations within an application. This expert response discusses some of the numerous ways to monitor and troubleshoot performance.

Virtualization platforms monitor virtual CPU calls well
Because the primary purpose of a virtualization layer or hypervisor is to create completely isolated partitions, it is often difficult to get a complete and accurate view of overall CPU utilization at the level of the host operating system or from within individual guest operating systems. Overall, most virtualization products do a good job of processing CPU calls. If you want to see what's going on under the hood, though, you'll need to use specific performance monitoring counters.

Virtual storage

Troubleshooting a SAN storage configuration in a Microsoft environment
Virtualization expert Anil Desai offers advice on troubleshooting a storage area network configuration in a Microsoft environment. While enterprise management tools may do the trick, our expert advises using virtualization-specific management applications in order to monitor the storage area networks.

Screencast: Storage management in virtual environments
Consolidating workloads onto virtual servers can increase performance, but a poorly executed strategy will only augment your storage management woes. This tip helps keep a strategy on track by outlining how to estimate capacity needs, assign workloads, select storage methods and monitor resources.

By default, most virtualization platforms will treat all your VMs equally. They'll all get the same priority levels and access to system resources such as CPU time. Microsoft Virtual Server provides two methods for managing the priorities of virtual machines: weight-based and constraint-based resource allocation. This screencast describes the two methods and provides a live demonstration of how to manage these resources on test and production host servers.

Analyzing statistics

I/O limit statistics for various virtualization platforms
Need help finding benchmark statistics for I/O in your virtualized environment? An expert offers some suggestions in this expert response. When looking at a disk, some folks may think a server with 100 or 200 I/O per second (IOPS) is high. Talk to an EMC rep about DMX storage and they laugh at 200 IOPS for an individual server. Then again, if you run an MSA 500 as the back end for a virtual machine platform several 200 IOPS virtual machines will start to show a performance hit. Again, it's still relative. A hypervisor will show only going to show limited overhead, but the hardware underneath has to handle it.

Virtualization Platforms

Microsoft Virtual Server and Hyper-V

Automate Microsoft Virtual Server to increasing server efficiency
Learn how to create scripts for performing routine tasks in Microsoft Virtual Server. When dealing with physical servers, many IT pros and end-users have found ways to use automated methods of handling otherwise tedious tasks. For example, performing backups and deploying operating system updates are generally done manually in only the smallest of environments. Fortunately, there are ways to automate the management of virtual machines in Virtual Server.

Managing virtual machine performance in Hyper-V with resource allocation
Hyper-V provides the same kinds of CPU resource allocation methods as Microsoft Virtual Server. These settings are defined per server, but with Hyper-V there's a slight difference if you use multiple virtual CPUs (which MSVS did not support). You can access the CPU settings in the properties of a VM, and the settings can be changed while the VM is running. This is useful for troubleshooting and quickly resolving issues when VMs get "stuck."

Advanced fine-tuning for optimizing Hyper-V performance
Hyper-V integration services and Microsoft's new synthetic network drivers require administrators to be proactive with hardware and network settings. This tip outlines how to fine-tune these settings to improve network and multicore processor performance.

VMware

Monitor physical server performance with the VMmark benchmarking application
VMware's free application VMmark can help you create benchmarks to determine the quality of hardware performance. An expert discusses how to download and successfully implement this management tool.

Using VirtualCenter maps to display VMware Infrastructure relationships
The mapping feature in VMware VirtualCenter allows you to visually present your virtual infrastructure in a variety of ways. An expert explains how to better navigate your infrastructure.

Obtaining server health status in VMware ESX and VMware ESXi
VMware ESXi can easily provide you with information about a server's CPU, memory, storage and more. But getting this info from ESX Server requires some additional steps. Obtaining server health status with VMware ESXi Server is easy. To do so, just open your VMware Infrastructure Client (VI Client), select the ESXi server in the inventory, click on the Configuration tab, then on Health Status. The full version of ESX Server, on the other hand, is quite different. To obtain the same type of health status information as with ESXi's feature, you must install the Dell utilities in the service console.

How to obtain, secure, install, monitor and log in to VMware's MUI using Linux
Unlike the Windows version of VMware Server, the VMware Server Management User Interface (MUI) is a separate installation component in the Linux version. This tip details how to obtain the MUI, how to install it, how to secure and configure it, and how to log in.

Monitoring VMware VI3 and logging events with VirtualCenter
Learn how to deploy software that will monitor VirtualCenter and log all VirtualCenter events and tasks to the Windows event log and a Linux Syslog Server. Most VMware Infrastructure 3 (VI3) monitoring can be monitored using existing monitoring software such as the Windows Event Log, a Linux Syslog Server or a vendor-provided hardware monitoring solution. Wouldn't it be nice if there was some way for VirtualCenter to send its logs to a Windows Event Log or a Linux Syslog Server? This tip shows how to do so using the VI3 SDK.

Hyperic manages both VMware and other virtualization platforms
Hyperic HQ is a popular agent-based systems management tool that can handle both VMware and non-VMware systems. An expert covers installation and the metrics Hyperic offers an administrator. Once the host has the agent installed, you can install agents on virtual machines. The VMware-specific monitoring information covers a lot of VM- and host-specific functions on ESX hosts.

This was last published in August 2008

Dig Deeper on Virtual machine performance management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

-ADS BY GOOGLE

SearchVMware

SearchWindowsServer

SearchCloudComputing

SearchVirtualDesktop

SearchDataCenter

Close