VMware vCenter Operations Manager (Part 3) – In Use

Home Articles & Tutorials VMware ESX and vSphere Articles Performance Tuning
VMware vCenter Operations Manager (Part 3) – In Use
by David Davis [Published on 14 Nov. 2012 / Last Updated on 14 Nov. 2012]

17
In this third and final article in the series, I’ll show you how to make best use of this new vSphere infrastructure performance monitoring and capacity analysis tool that you’ve just installed.

If you would like to read the other parts in this article series please go to:

VMware vCenter Operations Manager (Part 1) – Introduction
VMware vCenter Operations Manager (Part 2) – Installation
Introduction

In my past two articles in this three part series covering VMware vCenter Operations Manager, I first provided an Introduction to vCenter Operations Manager and then I showed you how to Install vCenter Operations. In this third and final article in the series, I’ll show you how to make best use of this new vSphere infrastructure performance monitoring and capacity analysis tool that you’ve just installed. As you learned in my first article, vCenter Ops (vCOPS) is offers unique features and functions and has a lot to offer. Let me show you how.

How vCenter Operations Collects Data

No matter what monitoring and performance tool I look at, one of the first questions I ask is “how does this thing get its data?” I ask this because how a tool gathers its data controls what it can and can’t do for you. If it doesn’t have the data that it needs, it’s unlikely that it can help you. Thus, one of the first things I wanted to explain is how vCenter Ops gathers its data.

Here’s what you need to know about vCenter Ops data collection:

vCOPS collects several types of data for a single monitored vSphere resource
Each of these types of data collected is called an attribute
A metric is an attribute for a specific resource in the virtual infrastructure
For each of those metrics, multiple readings are taken over time and each of those readings is a value
Certain attributes are identified as being more important than others as they could indicate severe problems in the vSphere infrastructure. Those special groups of attributes are called KPIs, or key performance indicators.
A vSphere admin can create “super metrics” that might, for example, track the average free disk space for all Exchange servers in your vSphere infrastructure
All of these metrics are collected from your vCenter server’s database and are moved out, into a vCOPS database, running inside the vCOPS Analytics VM
Because vCOPS pulls data from the vCOPS database, it can start providing useful performance and capacity information the first day it is installed because it can use the historical data, imported from vCenter
Logging In To vCenter Ops

While I did mention how to access vCOPS in the previous article in this series, I want to make sure I make it clear in this article on using vCOPS. Here’s what you need to know:

vCOPS is a web-based application
You will always connect to it using the vCOPS “UI VM” IP address or DNS hostname
You can access it in two ways – through the vSphere Client or through your web-browser.
If you access it through the vSphere Client, you’ll find it under Home -> Solutions and Applications -> vCenter Operations Manager
My favorite way to use vCOPS is through the web-based interface (not through the vSphere Client) because of the greater screen real estate that a regular web-browser can give you
You’ll need to know your vCOPS web administrator username and password. The default is admin and the default password is admin, as well. However, when you ran the setup wizard, you changed the default password to something that you (hopefully) remember.
You can also access the vCOPS Linux-based command line by bringing up the VM console or by SSH’ing to the vCOPS “UI VM” IP address or hostname
Here’s what the web interface login looks like:
Figure 1: Logging into the vCOPS UI VM Web Interface

vCenter Ops Dashboards and Widgets

Efficiently using vCOPS is all about taking advantage of dashboards and widgets. What’s that, you ask?

Dashboards – depending on your vCOPS privileges, you can create your own dashboards and add, remove, or re-arrange your own widgets. The Dashboard is where you will start using vCOPS. You can think of the Dashboard as the “homepage” of vCOPS and your table of contents to help you to dig deep down into your vSphere infrastructure performance and capacity. Most resources you see on the dashboard are hyperlinks that you can click on to dig down into more details and information about those resources. Each of the badges (critical to know and explained in my first article in this series) are hyperlinks that allow you to find out why your vSphere infrastructure’s health, risk, or efficiency are scored the way that they are scored. The dashboard is all about giving you a quick overview to tell you what if you need to take action NOW to solve a problem in the vSphere infrastructure or if you can go back to sitting by the pool.
Figure 2: vCOPS Dashboard Home Screen

Widgets – widgets are the small vCOPS “apps” on almost every page of the vCOPS interface which show graphs and calculate things like health, efficiency, time remaining, and reclaimable waste. If you have the right permissions you can configure the widgets, add them, remove them, or rearrange them.
Now, let’s get started monitoring your vSphere infrastructure with vCOPS.

Monitoring vSphere with vCenter Ops “Operations” Tab

Day to day performance monitoring of your vSphere infrastructure is done from the Operations tab. No matter what level of your vSphere infrastructure hierarchy you are at, the Operations tab will have five sub-tabs that break down into Environment, Scoreboard, Details, Events, and All Metrics. While the labels of these different sub-tabs are somewhat intuitive, the one that strikes me as most unique and useful is the Details. Here are the Details for my primary virtual datacenter:
Figure 3: Operations Tab Details for My Virtual Data Center

From this details tab, at the high view of my virtual datacenter, I can tell that I am “memory bound” but my health is 96, my workload is 22, and, so far, I have no anomalies or faults. By clicking on the workload badge, I can drill down to find out where this workload comes from and what resource is being pushed most, by what object (ie: “VM A” is utilizing the most memory and storage while “VM B” is utilizing the most storage I/O).

vSphere Capacity Planning with vCenter Ops “Planning” Tab

If your daily operating tab indicates that your vSphere infrastructure is running smoothly, what is expected to happen in the long term? Now move on to the vCOPS Planning tab. This is where you analyze capacity, for the long-term. After all, you need to know answers to questions like “how long before I run out of memory?” or “how many more virtual machines can I add on my current SAN?”

Under capacity, you’ll see the following sub-badges:

Risk
Time
Capacity
Stress
Efficiency
Waste
Density
All of these badges indicate the long-term capacity, from a specific aspect, of your virtual infrastructure. You’ll use these to find waste in your storage, how long before you run out of capacity of a certain resource, the risk of poor performance, and the stress that your vSphere infrastructure is under.
Figure 4: vCOPS Capacity Planning Summary Tab

In the above graphic, you can see that, at least in my lab environment, I have greater than one year of capacity remaining for CPU and can add 57 more virtual machines before I am out of capacity. It would be fascinating to look at similar capacity planning reports on production virtual infrastructures.

From the information gleaned from the daily Operations tab and the long-term Planning tab, you can both solve your daily performance problems and prevent long-term capacity issues in the vSphere infrastructure.

My Take on vCenter Operations

Certainly vCenter Operations is an impressive tool however, I have used 10+ different vSphere performance and capacity tools in my time working with vSphere. Every tool has its own set of good and bad. Here’s my take on vCOPS:

It’s pretty and the graphic interface is attractive
It’s for the enterprise, large scale, not for the SMB to quickly answer some performance questions
It can be complex until you understand the variety of badges (make sure you take time to recognize these quickly and use them to their advantage)
It’s extensible and can be connected to other systems like HP Openview
It’s big and will need a fair amount of resources, just to be powered on (the CPU and RAM requirements are in my first article)
It’s useful and can help you solve real problems in your vSphere infrastructure, potentially paying for itself very quickly
Don’t just take my word for it, try out vCenter Operations Manager for yourself by downloading the vCenter Operations Manager 60-day evaluation here.

If you would like to read the other parts in this article series please go to:

VMware vCenter Operations Manager (Part 1) – Introduction
VMware vCenter Operations Manager (Part 2) – Installation