VMware Aria Operations Reviews

Currently, the most valuable features of the product are monitoring, capacity management and planning; things of that nature.

We run a SCCM and SCOM. We took all of that and piped it in the back-end. That took a good amount of time; a couple of months to get it cleaned up and working. Once we had that completed, we first used it for capacity management and planning. With that, we were looking at what's over-provisioned, under-provisioned, things of that nature.

Moving forward, we took it to phase two, which is now. We're trying to do more proactive capacity management planning; look at forecasting on disk space, things of that nature. Now, moving forward, we're actually trying to move to a platform where we're going to try to make this our main monitoring base, too. We're going to build out portals. We're mapping everything as a service now, trying to go from individual VMs; we're trying to build everything out as a platform service and then build out portals so that we can publish all of these portals.

The benefits for us, because we're mostly virtualized, it's getting everything under one hood. That's probably the biggest thing for us.

It has most definitely helped us avoid outages or shortened outage time. I think that with forecasting and disk space features, we've easily avoided outages for machines. For example, we had older machines that had BDE partitions on them. With those, even if they're thin-provisioned, you had to take them down and remove that BDE partition before you could re-size them. With that, if you find something that ran out of space quickly, because we have a lot of growth there, we were able to forecast it, and say, "This thing is going to run out in three weeks; let's schedule some down time and get it cleaned up. Get some space added."

We use capacity management a lot and it's been pretty accurate for us.

With performance management, I think that some of the recommendations are a little off, but overall, it's been pretty good. They'll tell you that it's over- or under-provisioned. We've found that when we try to clean up and reclaim some resources, that might not necessarily be the case. Overall, we'll take about 50 to 60 percent of what they're saying we can remove and do that.

Adding it seems a little better with regards to saying that it's under-provisioned. We've found that when it's under-provisioned and we add the resources, what it's telling us is pretty accurate.

We're just getting into the new version 6 features now; more automation, increased integration with DRS for workload balancing and scheduling. We're still behind on some of it. We have not gotten into it yet.

We'd like to see more customizable things, more management packs; the ability to not have to customize the portals and do everything so ad-hoc. If they built more frames and shells into it so that you could deploy things easier and get it built out easier.

For example, and I'm not the primary one doing this, when you're building out the management monitoring portals and piping SCOM in and things of that nature, everything seems to be fully customized. There's no easy way to do that type of stuff. It should automatically be customized, or there should be templates or shells that you could use.

I'd like to see templates and other features built in, for when you're building out a portal and you want to give a portal and map out of all of your objects and services, and not machines themselves. I feel like that should either be built in or cleaned up so that you could build it in.

The UI can be a little laggy, at times; improving that would be nice. It just seems slow when it's loading out.

The organizational layout of it is pretty bad. There's a lot of information and a lot of tabs. When you're going to try and rifle through everything, it's very convoluted.

I have been using it for a year and half or two years.

We've had no stability issues to date. None. I know that it sounds crazy, but we did take it slow though.

Scalability has been good so far. We don't have a huge undertaking on it yet. Right now, we're using it for a couple hundred VMs and then maybe 300 or 400 VDI solutions. We're just starting the VDI side of things.

We have used technical support once or twice, and it hasn't been great; slow, a little inaccurate. We've worked through it; we're able to get the end result. It just wasn't as quick as calling in a BCS ticket. They were knowledgeable, and pretty good. It was just slow getting to the end of what we wanted to get to with resolutions.

We looked at it, we liked it. We talked to our TAM and they pretty much talked us into it. That's pretty much how we went with it.

I was somewhat involved in the initial setup. I have a main guy that does it. I was overseeing the project. I know that initial setup was fairly complex, but I don't think that it was ridiculous.

We looked at a few other vendors. It wasn't a very large offering. Also, for the price, it was very good. It was a very good price, we thought. We're educational too, so there's a different spin on that, as far as looking at third-party vendors versus this, plus we're trying to semi-unify on platforms and management. Trying not to keep putting more and more layers into everything.

Give yourself enough time to do it. It's going to take a little while. It took us a good six months to get it off the ground and functional. Probably another three to six months to get into the advanced features of it.

The most valuable features are performance troubleshooting and performance monitoring of VM CPU, storage latency, storage throughput, basic uptime, downtime.

A lot of the built-in free stuff that comes with VMware for troubleshooting latency has been valuable. When a VM is responding poorly, you've ruled out the basics – memory, CPU – and you're starting to look at the disk. We'll look at response time on the disk and if we want more details or history, we'll go and bring up vRealize Operations, which reliably catches it.

I don't know. I'm probably not the best person to address this question because I don't use it heavily. We have it installed in several places, but it's not our go-to for a lot of our products, for a lot of what we do. I work for a big organization; lots of different people doing different things.

I'd want it better integrated with the core products, so you don't have to go to a separate site or interface to go and use it.

The interface is a little bit too much, all at once for me. With the colors – green, red, black – if you drill down, it's a little bit confusing to me.

Also, it generates too many false positives. For example, with predictive growth, the predictive trends aren't very valid, realistic.

It's stable. No issues.

It's a pretty big environment. It's had no issues so far.

I have not used technical support for this product.

I think we've previously used Quest Defender and we’ve used SolarWinds.

I haven't seen a compelling reason to switch away from using vROps. We're a relatively large company, so I think it was part of an agreement in which they purchased it. I wasn't involved in the decision or even knew much about it beforehand to pick one versus another.

Upgrading it from vCOPS to version 5 was pretty straightforward. It was pretty easy.

It's probably one of the better ones out there. Just put your time in to researching it and setting it up correctly with the right data.

• It's become a lot simpler.
• We've started using views a lot more.
• The plugins for managing the virtual desktops, specifically the PC-over-IP and the latency.

Often times, an end user complains why something's slow, and you can immediately look at the dashboard and see exactly why. You have green, yellow or red, and you’ll see that latency's high, or you have some storage latency, as well, that is going to cause some slowness that the users are complaining about.

Most companies are trying to do a lot more with less. So, instead of being proactive, they're being reactive. There are pieces in vROps that allow them to become more proactive. There's some automation into fixing and remediating that are built into the product.

Often times, we hand this off to the customer so I can't give any specifics on whether vROps has helped avoid outages or shortened outage time. I haven't seen anyone actually save on storage. A lot of people use it for trending analysis; they don't usually move workloads around and actually show savings. As far as performance management is concerned, I have not really seen things speed up.

I would like to see more automation in terms of the remediation, instead of it just being a monitoring tool. Obviously, you don't want it to be causing issues in your environment and maybe you turn on features as you get more comfortable, but if it's just there to monitor, what's the point? You could use other tools, but if you're trying to do more with less and you're spending a lot of money on rightsizing this and deploying it correctly, it'd be better if I'd actually get something more than just a nice dashboard.

One of the bugs right now is, we've had some issues with single sign-ons; it's a known issue. It seemed like a pretty easy one.

The stability of the product has gotten a lot better recently. It's a nice dashboard tool for management, but the actual admins aren't using it on a day-to-day basis. When they do log in, something's locked out and they have to restart the services. They've had some issues, but it's become a lot better. We're still seeing it here and there.

Technical support has actually been pretty good. Obviously, pretty much for any company, tier one support is going to be what it is, but if you give them the information they need, I've actually seen some people who know the product really well.

Then the deployment of it is not as simple as it should be sometimes. To do it correctly, it's very complex. When you do it right, it becomes a very useful tool but if you don't do it right, it just becomes another nagging monitoring tool. One of the hardest parts is rightsizing it. There are tools out there to do that. Each environment is unique and it's not as simple as some of the competitive products out there. If you don't tune it, you're going to see a ton of red.

We've seen some customers deploy things for remote offices and they know that they're going to have some kind of slowness in the product. Instead of fixing itself, the baseline's there and it's not as intelligent as it should be.

I know Turbonomic is one that comes up all the time. A lot of times, for monitoring, we ask companies what they're currently using: Are they using SolarWinds or are they using something else, and how can we help them?

The most valuable features of the product are probably the monitoring and the dashboarding; for me, just from a management perspective, the dashboarding. It shows information for everything from, operational-wise, a lot of our daily tasks, things that were manual that you'd have to go and look at in the dashboard. They don't have to manually go in and do it; vROps flags it. They can just look at the dashboard each time. For us, it's really about the performance metrics; a lot of operational, tactical things. Being the manager, I don't have to look at it every day.

The other piece that I use is the projections, being able to see when I'm going to run out of resources, features like that. We use it a lot for resource planning, around where we're going to be next month. Every other week, I look at it and I can see where we're at on utilization of the environment and then use that for load balancing and similar tasks.

It gives me the information that I need when I need it. It's that easy.

It has helped us avoid outages or reduced outage time. Everything we're doing is proactive. If we weren't using our dashboards for management, being able to see where the issues are ahead of time, there’d be outages.

Even though I can't put a number on it, it saves us a ton of time with capacity management. We actually use that part of it.

Regarding performance management, I don't think it really helps us speeding up. It helps us avoid bottlenecks and things of that nature.

I'd like to see more granular reporting; the ability to simplify the reporting and being able to create the reports. There's a lot there, but when you dive down in there, I’d like to be able to break it out different ways. It’s been a challenge being able to train my folks to be able to get me the information that I want, so they can build those reports out.

The reports are usable, readable for the most part, but it's difficult to get my guys to be able to build the reports correctly, based on what I'm looking for. It seems like my requirements and the data mart can only go so far.

As far as the UI – for me, personally, accessing it because I don’t access it a lot – I feel like the usability could be a little bit better. They could streamline that a little bit better.

I've never seen a problem with it. We had one issue during an upgrade. We went with a clean upgrade. We had to run both environments for a little while, but outside of that, it's stable. It works. We’ve run into a couple of bugs here and there, but VMware is pretty responsive in getting them fixed. The bugs were with the reporting; when you're getting down to certain metrics within the data, to tweak it the way you want it. A lot of times they didn't think about, that's the way we'd want it. Items like that. How granular you get the data.

I’ve never had a problem with scalability.

My team uses technical support on a regular basis. I've always had a good experience with VMware technical support. No issues.

We used to use Foglight for that and we didn't like that. It was difficult to maintain and it was buggy.

My team did everything from ground up. Initial setup had medium complexity. It's not intuitive initially, so you have to go through a level of training to be able to get where you want to go.

Make sure you get training. Do the proper training. You need to have your folks trained to be able to manage it on an on-going question.

The most valuable features are historic trending and showing outliers; being able to determine if something is operating differently than it has normally operated in the last six months or even the last week.

We don't use it as much as I'd like, mostly because it's a little too heavy for what we have time to do with. vROps works best when somebody has it open 24 hours a day, is sitting in front of it, actively monitoring the heat maps and everything that's going on, and has the time to adjust all of the policies, so that when your operation is normal, you don't have any alerts going on, and you don't have any heat maps going on. I don't have anywhere near that kind of time. I'm the administrator of not only our global vSphere environment, the 500 servers that operate in there, and all of our storage infrastructure that supports our MetroCluster. Actually trying to spend hours and hours defining policies to get others to shut off that I don't care about, is very cumbersome.

It already has so many features that I can't utilize, it's difficult for me to determine what’s missing.

For me, the biggest area is out-of-the-box ease of administration. There are a lot of features that are turned on constantly and a lot of metrics that they use that, instead of asking you what you'd like your baseline to be, there are assumptions that are made about what good baselines are. Then, you have to go back and change all of those baselines so that it works for your organization. It would be nicer for me if the process interviewed you when you first bring it up about what kind of metrics you'd like to see in the different clusters in your environment so that that first day when you turn it on, you're not flagged as 90% of your stuff is out of compliance.

We just went through the implementation of vROps 6.2. We were coming off of an older version. I've been working on it as much as I can for about the last three weeks, and we still have some 150-odd active alerts that I'm going to have to go write policies to shut those off. That's sort of the complication. There's nothing broken in our environment; these are not problem machines. It's just metrics that fall outside of what the vROps team thought would be optimal for our environment.

I'd certainly like to be able to get more than I'm getting out of it. It's not a limitation of what it can do, it's just a limitation of the cumbersomeness of the management. My company is mid-range-sized and there are huge companies that have multiple people that actually just work on vROps. That's not, I would think, the majority of vROps customers. I suspect there are more of us, more mid-range-sized companies. It's a valuable tool, but it does require a ton of administration for the size of the IT org that we have.

I have been using it for about four years now.

Historically, it's been pretty stable. The biggest complication I have with stability is the fact that the management packs are written by a third-party organization and they don't necessarily interact well with all of the different versions of vROps. We have Storage Analytics from EMC as one of our plugins, but vROps only supports very specific versions of Storage Analytics for very specific versions of VPLEX hardware. Those very specific plugins can only be run on very specific versions of vROps. What actually determines the version of vROps we're running is the firmware revision running on our VPLEX. It's sort of the tail wagging the dog; I can't upgrade because then I couldn’t monitor my VPLEX. That's been very frustrating, as far as, "I'd like to upgrade and do this and do that and get this feature in, but I can't, because my VPLEX hardware code is too old."

I haven't talked very much to the vROps support team at all. I've only had, I think, one open case, and I resolved that before they got back to me. He was very nice when I talked to him, but I haven't really overly utilized support.

We've used a lot of different products for a lot of different parts of what it does. We mostly used vCenter monitoring and vCenter alerts prior to using vROps.

I didn't do the original vCOPS installation.

I did do the 6.2 installation. The actual out-of-the-box installation's not terribly difficult. Some of the certificate pieces for getting that plugged in is a little more cumbersome than I'd like, but that's a global thing with VMware in general; the certificate management is not where it needs to be.

It's not something that out of the box you just throw in and is going to give you a good return on investment. You really are going to have to use it and have somebody managing it to get your money back.

Make sure that you have sufficient resources to manage vROps before you actually pay for it. It's a very expensive product in the mid-range for what you get for the cost; you're going to pay a significant amount. You need to make sure that you're willing to dedicate the resources to that from a salary perspective. You're going to have to hire a person that helps manage this, or get some resources or free up some resources to really help do that, or you're going to waste your capital expenditure on vROps.

The alerts are the most valuable feature of the product. We run terminal servers and I like the granular ability of it getting into the guest OS. It gives us alerts for the hard drive, how is it on space, or pretty much anything happening at the guest level. It gives us a single pane of glass.

We use it a lot for capacity planning. Like I’ve mentioned, we do terminal servers. What we've been doing lately is giving it a metric that shows how many more VM's can we fit on that particular host. It's been great for that, and told us, with the research that's available, we can fit 10 more VMs on that host. From a capacity planning view, it's been great.

It has definitely helped us avoid outages with our internal servers. We've been having an issue with temporary profiles filing up the hard drives, and early on, before we really had the monitoring, that terminal server would just go down. Now at least we have something that says we have 10% hard drive space left on this particular machine; we get the alert in vROps, and we are able to get to it before it goes down.

It has not been helpful from a capacity planning point, but we have done a lot with the oversized reports, tuning some VMs. We were able to pull back vCPUs, memory and storage. We can reclaim some of that space that was being wasted on oversized machines. I guess that's capacity planning to some degree, but tuning is more what we use it for.

We also haven't really used the performance management features too much.

Even though the set up was kind of straightforward, getting the lay of the land was at times kind of confusing; I don't know if that's an indictment on me or the software. There was documentation readily available, it just wasn't as intuitive as I thought it would be, straight out the box.

Other than that, I don't really know about specific areas with room for improvement because at this point, we are just scratching the surface of what it does, because we are a smaller shop. We just had a merger, so at the top of the year, we're going to be onloading, onboarding a lot of DRs and scaling pretty quickly. Then I'll be able to really take it for a test drive.

Stability-wise, I haven't had any issues.

Because we are still a smaller shop, we haven't really had to scale it. Pretty soon, we probably will, because we are going to add some VMware Horizon for our VDI and NSX. I'm pretty sure we're going to have to scale it.

I haven't used technical support for vROps.

We weren't previously using anything. Just having a central measurement point was light years ahead of trying to go into each one, or setting up different types of software that have monitoring. It was just easier doing it with vROps.

Initial setup was actually pretty straightforward. I pretty much launched the VM, went into the web query and set it up from there; set up all the sensors, took you through a wizard. That was pretty much it. I didn't have to read about any customization.

Before choosing this product, I did not evaluate other options. We actually bought vSphere with vROps, so we never even looked at anything else.

Definitely go with it, if you're looking for a centralized management point, as far as being able to monitor everything, without having to manually go in and get sizing reports for VMs. It even reports on networking issues.

I'm waiting to see when we go into VDI and NSX; I think we'll really open it up then with a lot more monitoring options.

The most important criteria when selecting a vendor like VMware is pretty much support. We've had some vendors where it's not that the product’s bad, but when it does go south, the support hasn't been there. We definitely research, whether it's going online, looking at message boards, just kind of getting a feel for what other customers’ expectations are, if the vendor is meeting them or not. That's one of the things that we are really big on, the support perspective.

The most valuable feature is the fact that it understands what is normal for a workload and alerts only if you go outside the given boundary. It knows if a workload spikes at the end of the month, that doesn't mean it has to alert, because if that workload is spiking every time at the end of the month, it knows this is normal for that workload and will not alert. It'll only alert if that spike goes beyond the normal range of that spike.

It has the ability to filter and alert you only when you want to be alerted. It understands the IO profile of the workload. It knows when it has spikes, when it has valleys, in a manner of speaking, and accordingly will alert you.

It's a great troubleshooting tool. If you have enough management packs in it, you can see the entire supply chain all the way from storage to compute. It helps you see where exactly, potentially the problem is happening.

Of the newest features, the workload balancing is something we might use. Currently, the way our clusters are laid out we truly do not have a need for that use case. We are not a good candidate for that use case, but the fact is it might help us when we try to consolidate data centers. It can help us to take two clusters into the data center and perhaps migrate workload between the two. It might be used for a data center migration, the way I look at it.

As far as using the capacity and performance management features to save on storage, I cannot answer to that because we have a different team for storage. They manage the storage, they monitor performance and capacity. We monitor the compute side. We use vROps on a regular basis from a capacity management for the compute side; the built-in features, views, heat maps, and whatever they have are pretty good indicators of when we need to add capacity. It has been pretty reliable from that perspective in the sense that it tells us we have a defined threshold that it takes us x number of days to add capacity. It has been pretty reliable from that perspective.

I haven't yet truly used it for proactive monitoring. It's been reactive, but it helps nail down the issue very quickly, based on a VM, a host, or whatever. Their views are the biggest source of views out there.

I think they need to make the UI a little bit more simplistic. It can be a little overwhelming for people who have never used the tool before. For someone who is using these products, you can find things very easily once you're in the UI, but we tried for our users so that they can go in and look for their stuff in there. If they can make the UI a little bit more simplistic, that would probably be one thing I would ask for.

We are trying to empower the users. They should be able to go in and look for their VMs and do minor-level troubleshooting and similar tasks. The UI is a little cluttered from that perspective. If they can make the UI a little bit easier, similar to Google, it would help a lot.

We run infrastructure. Users have a mindset of different things they look for. For them, if there's a custom dashboard that we could set up with a very simple basic UI, where they can see the obvious things. They could just jump in there and see that dashboard, see where the problems are happening right away, instead of moving all over the place. That's why I haven’t given the product a perfect rating.

If it is designed well, it is quite stable. You need to know how many VMs will be reporting up to it. Based on that, if you stand up the cluster with a sufficient number of nodes; data nodes, management nodes and remote collectors. If you design it accordingly, based on the requirements, it performs really well.

So far it's been working well. We have a pretty big cluster; it's around seven nodes. It has been working just fine.

I have run into issues and the support I have needed was more from a guidance perspective than any big help; just asking them, having an upgrade, what would be the steps? What is the recommended procedure, if any? Is there any good guidance around it? They have been pretty helpful with that.

I haven’t actually had many issues with technical support. Once they kind of laid it out, given the environment, “This is how we would recommend that you do the upgrade.” The upgrade itself takes time just because of the scale of the environment. Beyond that, not much.

I was not involved in the initial set up, but I have been involved since then. The solution was already stood up by another team member of ours. He's no longer there, so I inherited the solution, but I have expanded the cluster and I've incorporated additional BUs that we have all over the U.S. and they're reporting into vROps now.

It's a pretty big product. From our perspective, it does a lot for you. You just need to do your homework and try to understand what you're looking for. It has all the answers in there. It does. You just have to know what you're looking for and know where to go to look for it because it can be a very complex product for a first time user. It can be.

The alerting is probably the most valuable feature of the product. If issues come up, we can tell right away to look in the system, dive a little deeper and see if there's a real problem and if so, we can fix it before there's an outage.

I definitely think some of the adapters could be enhanced.

When building out metrics, it would be better to have a more granular approach to that. Sometimes the metrics are all or nothing, and it would be better if you could adjust those to make it fit your particular needs better. On occasion, we can't even use some of the metrics because of the way our environment is set up; I can't turn it off for this pool and then have it on for this pool. It will give errors and then those errors don't mean anything. If you give errors that don't mean anything, then people stop listening to them and then bad things could happen. I would definitely like finer control of some of the metrics and how they work.

I have been using it for about a year and a half now.

It is a stable product most of the time. Occasionally, the adapters will go down and I'll have to reboot them, but that's maybe once every few months.

We haven’t needed to scale it yet. We have about six nodes and it's been pretty solid.

Technical support depends on the product. The people that work on the vROps side are definitely a lot more knowledgeable than some of the other support people I've worked with. Anytime I put a ticket in for vROps, it usually gets solved a lot quicker than a ticket for vSphere, Horizon or something like that. They've done a good job training their employees on the vROps part.

We switched from another solution to vROps because we weren't really happy with the first product.

vROps was attractive because it was part of our licensing deal so we figured if we don't have to pay for it, why not use it? Also, it does integrate a lot better with both vSphere and Horizon. We're big Horizon customers, so that was one of the big driving forces.

When I choose a vendor to work with, my criteria are that the product needs to be stable and easy to manage, but still be able to customize it to fit our environment; then, definitely, regular feature updates and bug fixes.

Initial setup was pretty straightforward. The Horizon adapter had some interesting licensing part to it. You just had to follow the white paper but it was a little tricky at first. Once I got the hang of it, it was no big deal.

My advice would depend on what you're trying to do with it. Our main goal was more of a monitoring solution, but obviously it does well in analytics, so I would ask what you're trying to do for it and then I could probably go into some of the details on what features would benefit you and if that was something you found useful, then yeah, great.