We were yet to fully deploy it as it should be. We've yet to fully implement it. Mainly because we have almost the entire HP OpenView suite, so that's been saving our life for right now, but we've noticed the benefits. We've noticed the night-and-day difference from monitoring everything from the hypervisor into the VM as opposed to from the VM's point of view out, without the machine knowing that it's really virtualized, that it’s sharing resources. It believes it is in a physical environment and it owns direct access to the CPU and memory, but it doesn’t: it’s virtual, it’s sharing it and it's pretend. vROps takes that into account because it all goes through the API for VMware.
Improvements to My Organization
Definitely, the first win for us was being able to see a CPU wait time where people were still building VMs as if they were building virtual servers, requesting too many CPUs. Not enough memory and they were really shooting themselves in the foot because you can have as many CPUs as you want, but if you're not really using them all, you're sitting there waiting for those virtual slots to fill up before it actually goes to the physical CPU. So you're adding so much overhead.
There were a couple of application teams that were able to take the constructive criticism, per se, and brought down how many virtual CPUs they had and they noticed a huge performance gain. Being able to do that for the environment was a quick win for us.
Room for Improvement
When you migrate from vCOPS to vROps, it has this awesome API where it grabs all the data, everything you've collected, and it puts it into vROps and you don't really lose that much. Everything you've already collected gets moved over and copied over and you're good to go. However, if you are on vROps and you're migrating to a major version of vROps or a new architecture design – like we're trying to do because we're trying to size it correctly – it doesn't go from vROps to vROps. I believe they had mentioned they were going to do that in the later version, or try to, but that would be my biggest request, because we need to build it out correctly and then migrate all that data we've already collected for so many years.
Aside from that, I would say getting around, creating your own custom super-metrics and all of that: It might not be that it needs to be easier to do, but maybe more well-documented.
Definitely reporting is nice and maybe they could develop an easier UI to do your own custom reports. We're still using all of the out-of-the-box reports, which are great. They've helped us hit that 70% of requirements, but it would be nice to have a nicer UI. Hopefully something like HTML that I can just drag and drop and just play around as opposed to the current UI that I have, which is like a popup; you have to know the metric name, and then somehow click over and get the metric. You really have to know how they're doing it and what they call their metrics and what they call the groups of their metrics and all that to know how to do the report right.
Use of Solution
We actually got it back when it was vCOPS 5.2, so I would say we’ve been using it for about three years.
Unfortunately, it has not been a consistently stable solution, because we've never fully deployed it as it's supposed to be. If you go through their sizing guide, we need I believe three virtual appliances tiered and we're currently on one virtual appliance. We have to reboot it often and it's just because it's not sized correctly. That's on us. We haven't had the time. We haven't had the resources. It is a big appliance, one of the bigger appliances that we own, but it's mainly because of what we're monitoring. We're monitoring so many VMs, so many data stores, so many network paths, and all that goes into, I believe, VMware’s equation for how it should be sized.
I believe it will meet the company's needs going forward once we size it correctly. Definitely its internal high availability is very simple to configure. We haven't looked at the disaster recovery for it. Unfortunately, we haven't given it the love that it needs to get it up to the way it's supposed to be, but I believe it will meet the company's needs.
Customer Service and Technical Support
Technical support is excellent; never had an issue with support with VMware.
Try it out. Yeah. Just spin up.
We have access to the software. I don't know how easy it is for somebody else that doesn't have an account with VMware or doesn't have an existing contract with VMware, to get the software, but for me, my solution, for everything that I have questions about, spin it out. They're all virtuals; why not? Worst-case scenario, you erase it. Move on to the next one.