Five ways to improve virtual server architecture performance

Administrators often work hard to find ways to improve performance and improve operational efficiency to tune the server virtualization architecture. Therefore, it seems that some minor adjustments and settings can be used to help a complex virtualized environment achieve high-performance experience and fast response support for users.

This article provides five ways to improve the efficiency of server virtualization architecture. These configuration methods can be applied to all server virtualization environments, and while some of these practices seem to be only minor adjustments, they can have a significant impact on the overall performance of the virtual architecture.

First, improve virtual infrastructure performance


by consolidating file servers typically, a file server for administrators is a big problem. Moreover, when we consolidate physical file servers for virtualization, we simply transform and migrate them from the physical architecture to the virtual server system. This approach means that administrators lose the opportunity to rearrange, delete, integrate, and organize some potentially unstructured data. So let's consider a completely different integration approach to optimize file server performance. One way to accomplish

file server functionality by using a storage area network (SAN). Many storage systems can support the external services provided by the CIFS (Common Internet File System) protocol. Therefore, the storage terminal can be conveniently connected to the AD domain control server through the network. This approach also enables virtualization of all file servers.

Another way is by means of an integrated way to achieve cloud file server virtualization. In this way, you only need to store approximately 10% of the data in local file server storage, and up to 90% of the data can be archived into the cloud storage system. From a functional point of view, a virtual file server appears to be external as if all data is stored on the local network. Moreover, this virtual file server can be easily connected to the AD domain control server, and at the same time can fully support the NT file system. Therefore, the management of the virtual machine server in this way is also very convenient, and there is almost no difference between the physical server and the physical server.

last method is to replace the original physical file server by running the virtual machine. In this configuration mode, you can further consolidate server naming to reduce the number of file servers. Windows systems can support multiple server naming methods, as well as NetBIOS naming, so that users can connect to the server (for example, \\\\ServerName) with some more short and friendly naming.

many administrators prefer to use a large file server instead of a plurality of discrete small file server. Windows Servers can support NetBIOS naming by adding the OptionalNames key in the registry

"HKEY_LOCAL_MACHINE\\SYSTEM\\CurrentControlSet\\services\\LanmanServer\\Parameters".

Second, disable Microsoft's clustering capabilities to improve system performance


there is a very common behavior while achieving virtualization is a pair of lower MSCS cluster of physical servers. For virtual machines that cannot support VMware HA and FT Fault Tolerance, MSCS provides an application-level failover capability. But this type of deployment is relatively bloated for many virtual server systems.

in a VMware environment, MSCS is one of the few ways need raw disk mapping virtual disks to support the installation of configuration. If users need to use the MSCS feature, they must abandon the most common deployment method of storing data on the VMFS file system. Such a process also increases the difficulty of storage allocation during the installation process, and the application is less common.

Therefore, disabling MSCS clustering capabilities Keyihuode many potential benefits. Assuming that your Windows system purchases an enterprise license or even a more advanced version, you can even reduce the cost of software licensing. If the physical server under the MSCS cluster is migrated to a virtual machine running the standard operating system during migration, you can save a certain license fee on the operating system.

Third, to improve the management capability


Caozuoxitong virtual server architecture server virtualization architecture to improve the manageability of the system by using a variety of management tools and methods. This approach is very helpful for the performance improvement of the virtual architecture, which ultimately benefits users.

e.g., optimal deployment is divided into a plurality of different groups of users based on different set of configuration policies, while minimizing localized single user unique operational requirements. This allows for the rapid deployment of multiple virtual machines from a single virtual machine template.

enterprise-class systems management software can also assist with the deployment of virtual servers. Many virtualization environment administrators hate to perform some program installation operations, as well as mapping network disks and localized configurations.

Therefore, the domain group policy management can be successfully deployed in most small virtualized environment, while large-scale enterprise environments can use Microsoft System Center, Altiris and Novell ZENWorks such as terminal management to achieve. These tools enable application delivery and interoperability management of sub-virtual machines based on policies.

After four very effective planning many common operations virus scanning process


virtual server architecture in a non-virtualized environment, integrated into the virtual infrastructure instead Will become a difficult issue. For example, the very famous virtualization environment virus scanning problem. The strategy we set is to complete a comprehensive virus scan of all systems at 3 am every Sunday (assuming that this point in time will not become a peak application period for countries in other regions of the enterprise.) This approach can also be applied to physics. Server, because the processor is mostly in the Idle state during this time.

In addition, problems caused by the virus scanning process will be more prominent in the SAN environment because it will bring great pressure to the local processor resources. If we do defragmentation in the virtual server architecture, we can understand the enormous pressure on the entire processor resource pool when all the sub-virtual machines start running full scan operations at the same time.

To avoid this, you can set different scan times for different virtual machines. For example, a scan of 15 virtual machines is started at 3 am, while another 15 starts at 5 am. This setting method is very necessary. Only in this way, some advanced features, such as VMware's DRS, can automatically migrate workloads based on an understanding of the computing power of the cluster system.

V, as far as possible during the new virtual machine in lieu P2V conversion improve system performance


Although the physical to virtual (P2V) conversion is a very simple operation But often new systems are purer. In Windows systems, the P2V conversion process will migrate many legacy configuration information, such as hardware drivers, some unnecessary software, and confusing file systems, which will affect the performance of the system.

order for the new system to achieve optimal performance, the best way is to deploy a new virtual machine, and then re-distribute the software needed. This includes the reinstallation of user applications, the conversion of code and configuration information, and the rebuilding of database systems on the database server. Of course, P2V conversion is still a very useful feature, because many administrators do not have enough time to reinstall all systems.