What is our primary use case?
We operate a test lab and we do virtual and physical testing for various clients.
We frequently build out a test sweep that looks like the client's environment and then runs the tests on that. We particularly do it with upgrades and things of that nature.
vROps is used to do performance measurements. It's in conjunction with two other products. One of them is called vRealize Log Insight and the other one is called vRealize Network Insight. That gives us a reasonably good profile of the performance in one of the systems under test.
How has it helped my organization?
Our business is built around testing, measurement, and performance measurement and vROps is the primary tool. We use it in a VMware environment and we do tests in other environments. But in the VMware environment, vROps and the associated products, Log Insight and Network Insight are the primary tools that we use. It's a basic tool. It's very important for our organization.
vROps provides proactive monitoring up to a point. There are limitations on its visibility. We often use it in conjunction with an operating system-specific monitoring tool. vROps provides not bad visibility into operating systems such as Windows and Linux, but if you want to track down problems in those, you're probably looking for something that runs inside the operating system. vROps is very important for the availability of the test lab.
Surprisingly enough, VMs take much fewer resources than most people think. vROps has enabled us to run 30% to 50% higher in terms of density. A lot of the work that we do is testing workloads, so the process is basically setting up a workload, guessing what the infrastructure's support that workload is, driving a test workload into it, and then manipulating the infrastructure until it begins to break or slow down. vROps provides the monitoring that tells us when those breaks occur, primarily at the hypervisor level.
vROps has enabled us to replace multiple tools. The performance measurement suite from VMware is three basic tools, vROps, Log Insight, and Network Insight. We use that cluster of tools in preference to things like Splunk and various other tools that are out there. It's a core tool for what we do. It is our measurement instrumentation tool, so it's critical to what we do.
What is most valuable?
In engagements with clients, we will often use vRealize for operational monitoring and that sort of thing. But our facility is primarily a test lab, so we use it for profiling and performance measurement.
For people who know it, vROps is quite user-friendly. It takes a little while to come to grips with it because it has a reasonably complex interface. The newer ones have gotten better in terms of being able to declutter the interface, but even so, there's a lot on the page, particularly in a reasonably sized infrastructure.
We've only just started experimenting with Tanzu to learn how to use monitoring and management. I have worked with Tanzu with a client who's in the process of post-deployment work. But I haven't used vROps specifically with Tanzu.
vROps enabled us to be more proactive in anticipating and solving problems. This has decreased our mean time to resolution by 40% to 60%.
It's not a huge concern of ours but vROps' workload placement increased VM density.
We integrated vROps with vRealize Log Insight. It provides alerts, correlates metrics, and checks logs across all of the components of our infrastructure. When you're doing this, you get a slew of performance information that comes up in real-time on the vROps console and interface. Much of it comes through logs and Log Insight processes that are in the background and then push back the results from the log processing up to the vROps dashboard. It identifies issues that are showing up in the logs. The integration is very useful to the testing process.
vROps and Log Insight provide us the instrumentation that allows us to identify problems and issues and look at possible solutions.
What needs improvement?
Deployment is still a little bit of a nuisance but you only do that once.
If I put on the hat of a client, I would say cost needs improvement. For clients with reasonable-sized infrastructure farms, you're looking at licensing at either per socket or per VM, and if you have an installation of any size, you're doing it per socket, and the per-socket licensing is a little heavy. Per VM license, if they have large numbers of VM, it is just not practical.
For how long have I used the solution?
I have been using vROps for six years.
What do I think about the stability of the solution?
We found it to be very stable.
What do I think about the scalability of the solution?
Keeping in mind that we're a pretty small operation, scalability is not an issue for us. In larger data centers, my experience has been that it scales very well.
Half of an employee's time is needed for maintenance.
Because we're a testing unit, the workload that we have in terms of testing will expand with the business. Generally, we run it on around three or four platforms at any given time.
We'll expand usage as our business expands, as we have more requirements, but we don't have a plan that says two months from now, we're going to add some more.
How are customer service and technical support?
We didn't contact technical support for vROps. When we contacted support for other solutions, they generally provided reasonably good support. They tend to stick with the problem until it gets sorted out, and usually, they're good at identifying what information they need and how to get it to them. Working with them is reasonably good.
How was the initial setup?
We've done the setup a number of times, so from our standpoint, it was pretty straightforward. But for someone just starting out, you really have to spend a lot of time with the documentation and understand the various configuration parameters and how they affect the operation of it. The setup is reasonably complex for a client.
Overall, the quality of VMware's documentation tends to be fairly dreadful. And so, you do a lot of searching around and bouncing back and forth. One of the biggest improvements they could make would be to actually use illustrations in the document so that there is a straightforward way to understand what the documentation is trying to tell you. It's very verbose. Trying to relate what's in the documentation to what's in front of you doesn't always go well. Documentation doesn't seem to move as quickly as the interfaces.
We're certainly not a large data center by any sense of the word. We have about 20 hosts. If we were to do it starting from scratch and moving up, the setup would take about two weeks.
It takes two to three hours per host but there's a lot of carry-on between the time you spend working on the hosts. There's preparation and various other things. Overall, it takes around one to two hours per host.
When we started, we installed vROps, linked it to vCenter, picked a group of hosts, and set up monitoring on that group of hosts and on the VMs in that group of hosts. We worked out all of the kinks from the configuration and setup. Then from there, we just rolled it out to the rest of the hosts and set it up so that at the beginning of a test, we can deploy what we need for a given host. It's not just vROps, but it's also the support things that need to be in place for us to quickly turn around a testing environment.
What was our ROI?
Most clients see a good return on investment in reduced staff time, they get early warnings about problems that are coming along, reduced time to diagnose and come up with solutions to problems. In my mind, looking at our clients, the people who use it in production operations, there is a return on investment. It depends on the size of the organization and that sort of thing, but typically, I would say you get a 1.5:1 return on investment and perhaps a bit more. It's very client-specific. This is associated specifically with the testing work we do with VMware installations. We do work with other installations that use Microsoft and various other things, vROps is interesting, but not really that useful. There are better tools for those other environments.
What's my experience with pricing, setup cost, and licensing?
There are no additional costs to standard licensing. There's time, material, staff cost, but they are not out of line or unusual.
To really use vROps effectively, you have to have three of VMware's tools. vROps, Log Insight and Network Insight. I'm not sure that would apply to every customer, but certainly would for the kind of work that we do. In a sense, the additional costs are those additional products.
What other advice do I have?
When it comes to efficient workload placement, vROps works with vCenter for workload placement, and vCenter carries most of the burden for that, so I'm not sure that's something that vROps itself does.
If you're running an evaluation or testing on VMware environments, vROps is really the only tool that makes sense.
My advice would be to find a specialist.
vROps will point you to where to look for the problem. When you actually dig into doing diagnosis and so on, you really need a good log processing facility to be able to dig through the logs and identify where the problems have arisen. vROps will notify you of a problem and will point you to where the location is. But to get down and identify the problems, you really need the log processing part.
Against other products, I'd rate vROps a nine out of ten.
Which deployment model are you using for this solution?
Which version of this solution are you currently using?