官术网_书友最值得收藏!

Planning for installation

Before you start installing vRealize Operations, you need to do some planning and preparation. The main areas you need to look at are product compatibility, architecture, and sizing. Due diligence in these areas at this stage will make for a successful implementation of vRealize Operations.

Product compatibility

Ensuring you are installing the solution on compatible equipment, and that the systems you want to manage are compatible, is the first thing to check.

Note

This chapter refers to the installation of vRealize Operations 6.1. You can find the latest requirements in the release notes for the version you are installing at https://www.vmware.com/support/pubs/vrealize-operations-manager-pubs.html.

vRealize Operations is relatively flexible with respect to the interoperability of the older versions of vSphere, so I have rarely come across situations where compatibility with vSphere holds a project up.

The following is a summary of the compatibility requirements:

  • vSphere: vRealize Operations comes as a Virtual Appliance (vApp) and can be installed on an ESX/ESXi host at version 4.0 or later. The host needs to be managed by vCenter Server 4.0 Update 2 or later. vRealize Operations can connect to, and manage, ESX/ESXi 4.0 and vCenter Server 4.0 Update 2 or later.
  • Browser: You access vRealize Operations using a browser. Most versions of Firefox and Chrome are supported, and I generally recommend using those as they have provided me the best experience. Internet Explorer 10 or 11 is also supported, however, you will experience slower performance.

    From a practical perspective, you need to access vRealize Operations from a device with a good screen resolution and screen size. For lab work, my 15" laptop at 1440 × 900 resolution is about the minimum I would consider. For day to day use, you will benefit from a decent sized external monitor.

    Using a tablet is almost impossible unless you were to build dashboards for specific use cases.

  • VMware products: The following integrated products are supported:
    • vCenter Infrastructure Navigator 5.8 and above
    • vRealize Configuration Manager 5.6 and 5.7.x and 5.8
    • vRealize Operations Manager for Horizon 6.1.0
    • vRealize Hyperic 5.8.4.x

Architecture

The building block of vRealize Operations is the node. You will build your vRealize Operations cluster out of one or more nodes, depending on your scalability and high availability requirements. Your architecture will also have quite a high dependency on sizing, which is covered in the next section.

Sizing

Accurate sizing is very important. Sizing is determined by the number of objects you will be managing and the number of metrics you will be collecting. If you are just managing a vSphere environment, this is relatively straightforward to calculate; however, if you are going to be adding a number of Management Packs, you will need to do more calculations.

The sizing guidelines do change from time to time, and from version to version, so you should always check the following Knowledge Base article for the latest information and sizing spreadsheet: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2093783.

The preceding URL has links to a sizing Knowledge Base article for each version of vRealize Operations. Each of these articles has a sizing spreadsheet that can be downloaded.

The sizing spreadsheet is fairly straightforward and contains three worksheets.

Overall Scaling Table

This worksheet highlights how many objects and metrics each size of node can support. You need to ensure the nodes in your design are large enough to support the quantity of objects and metrics in your environment, and do not breach the maximums in the Overall Scaling Table.

Sizing Guide – basic

This is suitable for simple deployments against environments in which you are only going to do vSphere monitoring.

Sizing Guide – advanced

This should be used if you are using any additional Management Packs. Unfortunately, there is not much detail in the sizing spreadsheet on the quantity of metrics each Management Pack collects. This means that you will need to check the Management Pack documentation to find out how many metrics are collected for each object type that the Management Pack monitors, and multiply that by the number of each object type.

Scaling up and scaling out

The following screenshot shows a typical sizing example. You can see that with High Availability disabled, you would have a choice of deploying with two small nodes, one medium node, or one large node.

Generally speaking, I prefer to scale up before scaling out as it decreases the complexity of the solution, and means that you have fewer moving parts.

An exception to this rule would be if you needed a larger number of concurrent user sessions. These are limited to 4 per node, so if in the preceding example, I needed 8 concurrent user sessions, I would choose to deploy 2 small nodes instead of 1 medium node or 1 large node.

Tip

The limit on the number of concurrent users per node has relaxed slightly with the release of vRealize Operations 6.1. Further guidance is expected, particularly for large multi-node clusters. Guidance for each version of vRealize Operations is documented and updated at the sizing URL referenced earlier.

High Availability

The final consideration is High Availability (HA), which was initially discussed in Chapter 1, Introduction to vRealize Operations Manager.

The first node installed in a cluster is the Master node. When you implement an HA cluster, one of the nodes will become a Master Replica node to take over the Master role should the Master node fail.

When it comes to deployment, you should ensure that the Master and the Master Replica nodes are hosted on separate hardware to ensure continued operations in the event of a host failure.

Note

Due to the in-memory Gemfire based database, vRealize Operations nodes need to be closely coupled and generally located in the same data center. Check the sizing URL referenced earlier for network latency requirements between the nodes and the remote collectors.

So, once you have put your numbers into the sizing spreadsheet, worked out how many nodes you are going to need, and what size nodes you are going to install, it is time to install vRealize Operations.

主站蜘蛛池模板: 宁武县| 抚松县| 稷山县| 博乐市| 望江县| 新宁县| 庆元县| 龙游县| 大同市| 剑川县| 北宁市| 建昌县| 麻城市| 新乡县| 张家口市| 丽水市| 乌拉特后旗| 玉环县| 白玉县| 依安县| 和平县| 万州区| 临邑县| 离岛区| 寿宁县| 绥宁县| 皮山县| 昌黎县| 建宁县| 永康市| 汉寿县| 宁阳县| 化州市| 淮北市| 二连浩特市| 江源县| 临高县| 连平县| 黄山市| 怀集县| 中阳县|