Opvizor Performance Analyzer Review: Simple, Elegant, Efficient Monitoring for Your Virtual Environment
Opvizor’s Performance Analyzer is a performance monitoring and optimization product for virtual environments, with a specialization in VMware vSphere. Performance Analyzer presents a dashboard that is simultaneously comprehensive and easy to understand and use, but also extremely customizable. Let’s take a closer look at what it can do.
Performance Analyzer is most useful in an environment with at least 100 virtual machines (VMs). These types of settings often suffer from performance issues, including CPU and disk I/O bottlenecks. As virtual environments grow, the hosts and supporting resources start to become constrained as more and more virtual machines contend for resources.
Worse, through valiant attempts to correct apparent performance challenges, well-meaning virtualization administrators often address symptoms rather than the root cause. For example, if a virtual machine (VM) is running slower than expected, a virtualization admin might decide to add more vCPUs to that VM. Unfortunately, this very action can make the problem even worse, rather than improving the situation.
Properly identifying the source of performance problems is a critical first step in understanding how to fix things. Fortunately, Performance Analyzer makes this much easier through the product’s intuitive dashboard.
Performance Analyzer gathers and visualizes performance data from all your datacenter components, including hypervisors, storage, operating systems, containers and cloud providers. More information equates to faster identification and resolution of trouble spots. Without the ability to go beyond the hypervisor, a performance management tool simply can’t be effective.
Installation and Setup
Despite its comprehensive list of data sources, Performance Analyzer is easy to set up and use. It starts with installation, which is a breeze. Performance Analyzer is delivered to you as a prepackaged virtual appliance: just download the OVF template from Opvizor’s website at try.opvizor.com/perfanalyzer and install it. Opvizor grants you a 30-day free trial of Performance Analyzer that doesn’t carry any feature restrictions other than eventually timing out.
Once you’re up and running, import the virtual appliance using VMware vCenter or Microsoft Hyper-V Manager. Next, start the appliance, fire up the console, and open the appliance configuration page page under http://ip_or_systemname_appliance/admin in your browser. (Step-by-step instructions for this can be obtained by clicking here). Figure 1 shows the admin logon screen.
The tool will prompt you to provide a connection to your vCenter or other management software, and after just a few minutes while you wait for data collection to take place, you’ll be monitoring your entire virtual environment. Opvizor recommends that if you’re going to be monitoring more than 1,000 VMs, you use SSD for storage, as monitoring at that level is very I/O intensive. They also recommend doubling the amount of RAM for that kind of setup. It just makes sense that you’ll need something a bit beefier to monitor a larger or more substantial environment.
One of the first things you may notice (Figure 2) is how simple and unified the Performance Analyzer interface is. Data like total VMs, the physical CPU-to-virtual CPU ratio, host count, CPU usage and memory information is presented clearly. You can see at a glance, for instance, if there’s any memory ballooning or swapping going on. These are some of the key metrics to track to determine if your VMs are performing efficiently.
The bottom half of the main UI is shown in Figure 3. It shows the latency of your system and the total IOPS your VMs are consuming from storage. Again, Opvizor understands the most important information most virtualization admins want immediately, and provides it up front.
One of the goals behind Opvizor, according to CEO Dennis Zimmer, is simplicity. “People are tired of management tools’ complexity,” he says, and he’s right; especially when it comes to VMs where sprawl and resource usage become such big issues. In fact, one of the biggest hurdles faced with many management tools is eliminating the noise and getting the thing to focus on actionable intelligence.
One way Performance Analyzer demonstrates its ease of use is with outstanding informational popups that tell you exactly what a particular metric measures. But they go far beyond that: the popups often explain symptoms of a problem, what the specific problem is, and even how to remediate the issue. Just click on the little “i” in the top left corner of a panel, as shown in Figure 4. This is a tremendous help, as there are a lot of numbers to track. Further, it’s educational for the administrator. Eventually, learning about all of the metrics in a quick way ultimately streamlines future remediation efforts.
As you can see, Performance Analyzer comes pre-configured with a dashboard that meets the most common needs of many organizations. In fact, Opvizor indicates that about 60% of customers never need to make changes to Performance Analyzer. There are no thresholds to set out of the box, no changes to make for the majority of admins. If you’re looking for a no-brainer solution that you won’t have to mess with, something truly plug n’ play, this may be just what you want. This sets Performance Analyzer apart from much of the competition, which often requires a lot of effort to get to minimum usability.
Of course, if you do want to customize the product, that’s easy to do, and you can tweak it to your heart’s desire. You can change dashboards, and a create full-stack view or service-oriented view, for example, for more advanced monitoring needs.
You can easily drill down to get specific performance metrics, too. For instance, if you want to know which VMs are using the most resources, it’s just a matter of setting the right categories in the dashboard; you can instantly see the five (in this case) VMs using the most CPU, memory, read IOPS, write IOPS, etc. See Figure 5. This is incredibly handy when you need to find the big performance culprits with minimal effort.
Let’s take things a step further. What if the source of the problem you’re trying to track down is located in a SQL Server database? Simply open up the “SQL Server Monitoring” dashboard and have a look, as shown in Figure 6.
Similar to the main UI, at the top is a clear, well-thought-out interface with an overall report (in this case, it’s showing that the SQL databases are “HEALTHY,” with other pertinent information.) In the lower right-hand corner of each box is a little triangle icon which can be enlarged, if you want to take a more in-depth look at a particular metric.
Toward the bottom of the screen is a graphical representation of the database counters, as Figure 7 shows.
Hovering a mouse over any of the graphs triggers a popup that provides more data about the various database activities, without ever leaving the main interface. This is the definition of a “single pane of glass” monitoring solution.
You can create as many dashboards as you want for your datacenter. There’s a kind of “dashboard of dashboards” at the top, which allows you to search for a particular dashboard; it also displays a full list of the available dashboards, as shown in Figure 8. Just click on the one you want, and it appears in moments.
Very recently, Opvizor has updated Performance Analyzer to provide more out-of-the-box usefulness. The 4.8 release includes three new dashboards, which are highlighted in red in Figure 9:
- Spectre & Meltdown
- Starter: ESXi Host Troubleshooting
- Starter: VMware Virtual Machines
Spectre & Meltdown is especially interesting. It monitors your environment’s performance after installing the patch for that very serious Intel chip flaw, and can compare it to pre-patch performance. Since the patch is notorious for slowing down CPU and IOPS, it’s important to know its impact on your datacenter.
The “Starter: VMware Virtual Machines” dashboard (see Figure 10) checks the NUMA Home mode, a metric typically available only with the VMware command-line tool esxtop. This can tell you if NUMA locality is causing a system slowdown. As with the other metrics, it provides graphical clues as to where a performance problem may be lurking.
Opvizor does a fantastic job monitoring your servers, storage, virtual environment, and applications. It will monitor virtual networking and switches inside your virtual environment. However, it does not go beyond the confines of the application environment to perform full physical network monitoring. If you need a single solution that can do everything, including monitoring of physical networking hardware, Opvizor may not be what you’re looking for.
In such a case, though, you may want to consider rethinking your approach to monitoring. There are all-in-one monitoring solutions out there that can monitoring everything, but they’re often incredibly complex and incredibly expensive. The complexity alone can make these products non-starters. Stories abound about companies attempting to implement complex end-to-end monitoring, only to give up or to settle because the product is overly complicated or too noisy.
You may be better served by getting a tool like Opvizor, which excels at helping you support your workload environment, and a different tool that has laser focused on helping you support your network infrastructure.
Overall, Opvizor’s Performance Analyzer is a solid answer to one of the more growing problems in modern datacenters. As VMs proliferate, and storage, CPU and networking become more and more virtualized, keeping track of it all can be an admin’s worst nightmare.
Performance Analyzer can act as your aspirin. It’s a turnkey solution that understands the importance of simplicity and ease of use. It’s very true that the more complicated a monitoring solution is, the less likely it is to be used. Huge learning curves are something today’s admins don’t have time for, especially when solutions like Performance Analyzer are available.
Opvizor has a demo environment running on their website, so you can see how it all works: point your browser to pademo.opvizor.com. Once you’ve seen it, they’re confident you’ll want to get the free, 30-day trial, and then buy the product (pricing is not set, but rather done on a case-by-case basis. Their website indicates, though, that pricing starts at about $650 per year).
We give Opvizor high marks for Performance Analyzer and highly recommend that you put it through its paces for yourself.