Management and Optimization of VMware vSphere deployment (excerpt)

Source: Internet
Author: User

 

The following content is excerpted from the first chapter of VMware vSphere deployment management and optimization.

For A virtualization project, the first thing to determine is the architecture. Only after the selected architecture is selected can we design the next step and optimize the entire configuration. This part introduces three types of virtual infrastructure.

===== Excerpt start ====

In general, there are very few workloads that fully use Gigabit Ethernet bandwidth, let alone 10-Gigabit bandwidth. Therefore, this section mainly discusses network capacity planning from the perspective of the number of ports. How many ports does the virtualization solution need? To explore this issue in depth, we must first figure out three different virtual infrastructures: Traditional Gigabit Ethernet, integrated rack-mounted architecture, and integrated blade architecture.

Traditional Gigabit Ethernet Architecture

Figure 1.1 describes a traditional Gigabit Ethernet virtual infrastructure. Because it has been widely used for a long time, this is perhaps the easiest to understand architecture. Depending on the Storage Architecture, a typical vSphere host will have 8 to 10 Gigabit Ethernet ports connected to the upstream switch. Because we need to isolate different types of network traffic on the same vSphere host, this is a common best practice.

For example, the method for isolating traffic shown in 1.2 is quite common.

650) this. width = 650; "title =" Figure1.1 _ thumb "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; margin: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" Figure1.1 _ thumb "src =" http://www.bkjia.com/uploads/allimg/131227/1S15Q354-0.png "width =" 648 "height =" 536 "/>

Figure 1.1 traditional Gigabit Ethernet Architecture
  • Management Network-VMkernel2 NICS)
  • VMotion -- VMkernel2 NICS)
  • Fault Tolerance-VMkernel2 NICS)
  • Storage-VMkernel2 NICS)
  • Virtual Machine traffic-two virtual machine network adapters)
650) this. width = 650; "title =" cn-Figure1.2_thumb [1] "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; margin: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" cn-Figure1.2_thumb [1] "src =" http://www.bkjia.com/uploads/allimg/131227/1S15W926-1.png "width =" 462 "height =" 637 "/>
Figure 1.2 port group layout of traditional architecture

As the infrastructure continues to expand, more and more Network Ports and network cables will inevitably cause you trouble. Taking traffic isolation of 1.2 as an example, each host has 10 Network Ports and is connected to multiple switches in the network. This means that a 24-port switch can only meet half of the number of ports required by four vSphere hosts. When there are more and more hosts, consider other solutions that provide better scalability.

Integrated rack Architecture

Figure 1.3 describes an integrated rack-mounted architecture that integrates traditional Gigabit Ethernet connections into a 10 thousand Gigabit Ethernet infrastructure by connecting to an FCoE switch or a 10 Gigabit Switch. This allows you to expand your infrastructure while providing more bandwidth and reducing the number of ports required. In this architecture, the number of ports is more manageable. However, the number of cables and ports will continue to grow linearly with the increase of hosts, which means that the number of ports may still be a problem in a larger environment. However, you will have more room than the traditional Gigabit Ethernet architecture, without the need to add more switches.

650) this. width = 650; "title =" Figure1.3 _ thumb "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; margin: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" Figure1.3 _ thumb "src =" http://www.bkjia.com/uploads/allimg/131227/1S15Ra5-2.png "width =" 479 "height =" 602 "/>

Figure 1.3 integrated rack Architecture

Note:

FCoE is a protocol that allows encapsulation of FC data loads and transmission over Ethernet. This allows FC traffic to be transmitted over the same cable as Ethernet traffic.

Whether it is a converged rack-mounted architecture or a blade architecture, a vSphere host usually has only two Network Ports. Generally, we create a vSwitch and isolate different types of network traffic in the port group, as shown in Figure 1.4.

650) this. width = 650; "title =" cn-Figure1.4_thumb [1] "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; margin: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" cn-Figure1.4_thumb [1] "src =" http://www.bkjia.com/uploads/allimg/131227/1S15U202-3.png "width =" 509 "height =" 634 "/>

Figure 1.4 port group layout of Integrated rack and blade Architectures


Integrated blade Architecture


Figure 1.5 is an integrated blade architecture that uses UCSCisco Unified Computing System as an example. Note that, from the network perspective, the bandwidth and port count calculations occur on the Blade Mgmt in the Fabric interconnect device diagram of the UCS 6100 series. You can think of them as an aggregation layer of network and storage. Would you like to extend the downlink from the 6100 series to the blade server to 20 Gb? No problem! You only need to add two 10-Gigabit uplinks to each fabric interconnect device of the 6100 series. Want to expand to 40 Gb? No problem! Add four 10-Gigabit uplink links, and so on ...... In this architecture, you do not need to worry about the calculation of network bandwidth and the number of Network Ports, whether it is blade servers or network bandwidth. Of course, you still need to do some wiring work for fabric interconnect devices of the 6100 series, but the workload will be greatly reduced because the blade server can share the bandwidth. This architecture makes the relationship between the network and the vSphere host in the given infrastructure no longer meet the requirements of linear growth, but provides predictable performance and progressive network scalability.

650) this. width = 650; "title =" Figure1.5 _ thumb "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; margin: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" Figure1.5 _ thumb "src =" http://www.bkjia.com/uploads/allimg/131227/1S15R2A-4.png "width =" 579 "height =" 383 "/>

Figure 1.5 integrated blade Architecture

 

............
............

From the resource perspective, hardware selection is determined by capacity planning. However, management factors may also be included in decision-making considerations. Before making any hardware choices, you should first explore the traditional Gigabit Ethernet, converged rack mounting, and integrated blade architectures of different infrastructure architectures to understand the advantages and disadvantages of each architecture. Table 1.1 lists some discussions that are helpful for your judgment.

650) this. width = 650; "title =" cn_table1.20.thumb [1] "style =" border-left-0px; border-right-width: 0px; background-image: none; border-bottom-width: 0px; padding-top: 0px; padding-left: 0px; padding-right: 0px; border-top-width: 0px "border =" 0 "alt =" cn_table1.20.thumb [1] "src =" http://www.bkjia.com/uploads/allimg/131227/1S15V540-5.png "width =" 595 "height =" 196 "/>

This article is from the "delxu Tech Notebook" blog, please be sure to keep this source http://delxu.blog.51cto.com/975660/1195200

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.