Shulou(Shulou.com)05/31 Report--

How to carry out the principle analysis of vds and vcenter, many novices are not very clear about this, in order to help you solve this problem, the following editor will explain in detail for you, people with this need can come to learn, I hope you can gain something.

In the virtual network built by vmware, we often encounter the terms VSS (standard virtual switch) and VDS (virtual distributed switch). These are two more important and confusing nouns. After installing exsi on the physical server, the system defaults to deploying VSS, the standard virtual switch, to the virtual machine. A standard virtual switch can only be applied to one physical host, does not have any flexibility, and cannot share a virtual switch with other hosts. For the network structure with a small number of EXSI hosts and simple management, VSS has met the requirements. However, when there are a large number of EXSI hosts in the network, and several virtual machines are needed on each EXSI host, the maintenance work is also greatly increased, and VSS alone can not meet the actual needs. Compared with VSS, VDS can centrally manage all virtual machine traffic, and can achieve VLAN partition, NOIC-based security, flow control and other functions.

Let's talk about VDS in detail, and combined with VCENTER and other scenarios, let's take a look at the position and role of VDS in the overall virtualization architecture.

1. The logical composition of VDS

VDS can be divided into two logical parts, namely the data plane (data plane) and the management plane (management plane). The data plane performs actual packet exchange, filtering, and tagging. The management plane allows administrators to configure different parameters of the data plane function. As for the vSphere standard switch (VSS), the data plane and management plane appear on each standard switch. With this design, the administrator can configure and maintain each VSS one by one.

The data plane is the place where actual packet exchange, filtering, and marking operations are performed, while the management plane allows administrators to configure the functions of the data plane.

VDS treats virtual networks as aggregate resources. The single host-level virtual network parameters are extracted as part of a single large VDS that spans multiple hosts at the data center level. According to this design, the data plane is still local to each host, but the management plane is centralized, and vCenter Server acts as a central control point for all parameter configuration and virtual network management. Each vCenter Server instance can support a maximum of 128 VDS, and each VDS can connect up to 500 hosts.

Many concepts of configuring and managing standard switches are implemented by VDS, and some virtual network parameters have been changed to support extraction operations.

Distributed virtual port groups (DV Port Groups) are VDS-related port groups that specify configurations that apply to a set of distributed virtual ports. The distributed virtual port group also defines how to connect to the network through VDS. The configuration parameters of distributed virtual port groups are similar to those of port groups on standard switches, except that some advanced features have additional options. Virtual LAN number (VLAN ID), traffic × × parameter, port security, Nic aggregation and load balancing configuration and other settings are all configured here. Each VDS supports up to 10000 static port groups.

Distributed virtual uplink (dvUplinks) is a new concept introduced with VDS. Distributed virtual uplinks provide extraction levels for physical network cards (vmnics) on each host. The network card aggregation, load balancing and failover strategies on the distributed virtual port group are applied to the distributed virtual uplink rather than to the physical network card on a single host. When a host is added to the VDS, each physical Nic on that host is mapped to a distributed virtual uplink, thus achieving consistency between Nic aggregation and failover, regardless of how the physical Nic is allocated.

2. VDS and VCENTER

VDS is a distributed switch built across Host, so it is related to vCenter and can only be created under vCenter, not across vCenter. Because the underlying communication of VDS still needs a physical network card, VDS can be regarded as a unified switch with data plane and management plane.

3. The two port of the physical network card are regarded as two physical network cards and have two uplink links.

4. For the consideration of redundancy, two network cards can be satisfied. However, for all business, migration, management needs, depends on the quantity of these parts, so two or four pieces, is not a definite number. Maybe it will change over time. However, according to the current amount of data, two 10GE network cards are sufficient. Other factors need to be considered, such as whether your management network segment is a physically isolated out-of-band management network. If so, you need to use a separate physical Nic to connect to an independent physical switch.

Is it helpful for you to read the above content? If you want to know more about the relevant knowledge or read more related articles, please follow the industry information channel, thank you for your support.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.