Please share with the community what you think needs improvement with NETSCOUT nGeniusONE.
What are its weaknesses? What would you like to see changed in a future version?
The initial deployment is tedious and requires a lot of build, deployment and configuration time. Experience is key to a successful deployment. It is critical to work with the network team to understand how the data flows across the estate to ensure that you tap, aggregate, de-capsulate, de-crypt, and de-duplicate correctly. From these points, you need to ensure that the logical strategy and configuration suits the correct reporting and visibility as to ensure that you get the most out of any potential logical context. It is also important to ensure that you spend time to understand the applications you want to profile and work with the application owners to ensure that everything is aligned correctly.
I would like to see nGeniusPULSE and AED (Arbor Edge Defense) integrated with the solution. Some help screens would be nice, especially if we bring on new operators. It would be great to see if they have more helpful tips available. Between the host analysis and session analysis, there is a limitation of one hour. This can hinder us if a situation occurred ten hours ago. Sometimes, you can't open up that window.
It is a good product with a few limitations. It is so complex and takes a bit of training to figure out. We need better training, so we can take this complex solution and implement it more easily. Change the font size on the grid in nGeniusONE so the names of the grids will all fit on the grid tiles. The font is so large that you can't see the name.
They can improve the UI. For example, with all modern tools, they generate a shared URL, like a Slack URL. Somebody clicks and they see the exact same thing as you. With this tool, if you want to tell somebody how to get to your view, you have to give pointer steps. The single pane of glass is a decent effort, but it is not how things are done these days. It is not a good monitoring tool. It is more like response tool for us.
The product was lacking for awhile when they did the Arbor acquisition. I was waiting to see more security stuff, which they did eventually add, and is now impressive.
The single pane of glass is possibly overrated. The dependency mapping is good, but I am hopeful that they will build some type of partnership and relationship with ServiceNow. I want to see NETSCOUT partner with ServiceNow so they can leverage Service Now Discovery and Service Mapping to automate the build of the service dependency mappings inside of nGeniusONE. The user interface needs some updates. There is some complexity to the product. You have to understand where the InfiniStreams are and what physical interfaces are connected where, so when you go into the user interface, you know what data you are collecting and from where.
The single pane of glass view is a challenge. I like the graphics, they're easy to understand, but when more digging is required, it's more complicated to get what I'm expecting. Since the challenge for me is the dashboard, I would appreciate having a better view from the dashboard. What I don't know is whether the issue is that our configuration needs work. We probably don't do the mapping and the dependency configuration properly and that may be the reason why my dashboard is not crystal clear.
In terms of additional features, Bruce Kelly was talking about the NFV and 5G aspects of it, in monitoring all the APIs for all of those functions. We're really looking forward to seeing that so that we can give better visibility into the functioning of the cloud and the orchestrator itself. There is room for improvement in its stability and by expanding into the cloud and orchestration sphere, which I think is on the roadmap.
I need more details on the vSTREAM and how that scales from a CPU perspective. I know that we can start with one virtual CPU, but at the same time, our clouds are still limited by compute nodes. That's an ongoing question and it's part of why we're here at NETSCOUT Engage 2019, to see how we architect that out. I'd like to see improvement in scalability and the CPU perspective on the actual cloud nodes. It would be good to have a roadmap of what impact to our underlying cloud we will see as we add vSTREAM vCPUs.
There is a lot of the VoLTE, voice, video MOS, and customer experience that we'd like to do. There's a lot of throughput analysis where we're trying to understand, with the vendor, whether it's accurate or we need more work on it. Those are our top priorities. NG1 has been stable for a while in our environment - at least we have what we needed. But with nBA, there's a lot of room for improvement.
In terms of the single pane of glass view, it's good, but trying to set up dashboards is hard to figure out at times, if you don't do it every day. It's not really intuitive to set them all up. Other than that, it's a good dashboard. A lot of people are using it. If there were a wizard to take us through, step-by-step, creating dashboards and the like, that would be really helpful.
Some of the filters could be easier to see and to set up. That's the only thing that I've ever had any trouble with. The ones that I've seen here, at NETSCOUT Engage 2019, are part of a newer version that we don't have yet, and it looks better. So, it may already have been fixed.
We would like more encryption of customer data, because we have a very security conscious company. We have a lot of regulation coming in which requires us to make customer data private. There is a lot more integration work that needs to be finalized for simpler tool set. The integration of many products into a much simpler single pane of glass is where we want to be.
In terms of the single pane of glass view, NETSCOUT has many tools. I think NG1 provides that single pane for a lot of things, but users are still using several different applications within the NETSCOUT realm of applications, and more integration would be helpful. Our biggest area of concern right now, supporting the applications, is that while NETSCOUT does a good job of monitoring the network and the applications, we need more visibility into system health and performance monitoring. We need something that will monitor the tool that monitors the network.
The product is a little complicated.
In terms of additional features, they have the virtual clients here at NETSCOUT Engage 2019, and they have really expanded that. That type of coverage is going to be crucial. The COTS that they are doing now are a very good idea, to lower the price some. We work with them weekly, and if we uncover something, a feature that would be relevant, we usually report it. A lot of times it will get included. Regarding room for improvement, on a network the size of ours, the loading times seem a little extensive, 20 or 30 seconds to load up some graphs. But there is a lot of data being crunched. That's all server hardware.
In terms of the single pane of glass view, when we build it out in the nGeniusOne platform, there are multiple tiles and, depending on what we're trying to examine, it doesn't all fit in one single pane of glass. It would be nice to have that functionality, but you really do have to categorize things because there is so much data. The biggest thing is being able to provide net path. One of the products we use is SolarWinds, and it provides a very cool mapping of an agent from end-to-end. If NETSCOUT could somehow implement that into their design, whether it be sniffer-to-sniffer, or that kind of thing. I know they have some functionality along those lines, but if they could make it quicker and easier to get those net paths, it would be huge. I could quickly plug in problem IPs and get a full hosted view of where it's going from end-to-end. That would be really useful. Finally, the GUI, the interface, has room for improvement. It's user-friendly to a degree, but when comparing it to other products, such as in the Cisco environment or SolarWinds, I found that I could just fumble my way through those tools very easily without training. Whereas with NETSCOUT, I need training in order to set stuff up because I would never figure that out on my own.
While it is good, the single pane of glass view is too high level. It is better for management or someone doing sanity checks. A lot of times, I need to go deeper into the additional screens to get what I want out of it. We would like to have increased performance in the future. Eventually, we will need more horsepower. We would like NETSCOUT to add additional topics to the data that it collects, because big data is important going forward.
I like the Dependency Mapping the solution provides. I wish there was a better way to show large groups, greater than 500, instead of just not displaying anything. I would like to see it closer to more of an APM-type, or at least have that availability to compete with APM - the AppDynamics and solutions like that. I feel it's a natural step to at least have that available. Also, some integrations with ticketing systems like ServiceNow would be helpful.
We would like better end-to-end data flows. This is something that my users always complain about, as they don't know what the data flows are on the network. We have things, like Cisco NetFlow, but all you have is the two endpoints that you have captured from a probe. We would like to know every point along the line.
I'd like to see more data expert capabilities. That's one of the big things we're looking forward to, with the release of the KAFKA exports, and expanding the kind of data we can both import and export with the system. Also, for individual subscriber tracings, sometimes it does not capture all the messages. There is a little bit of room for improvement there.
This is not so much application-specific but rather is about the user experience: How the user experience has degraded is what I would like to see more in the tool. A lot of tools highlight what's going on but they don't actually pinpoint the user experience. It would be good if there were a small message or something highlighting what the user experience is like and any degradation that's actually occurring.
I would love to have them reassemble fragmented packets. That would be a very big plus in my book. While it does give me increased visibility while conducting IT deployments, I have experience some limitations with it. We have jumbo frames, which can get fragmented. They don't bring together everything that I need for me to work right. If I could get an end-to-end, then it would give me a good view of how everything would be in my particular call flow.
The feature I am looking for is the Arbor technology, especially to run in parallel with our firewall. That's one thing that I've definitely wanted and, eventually, it's getting there.
I would like more in-depth convergence between all the applications, especially when I look for information through a data mine. It has a lot of what I like to use, but some features are not there yet. It is sometimes even going down to older protocols still getting used in the world right now. They need to improve using voice other SIP.
The areas it covers are so vast that the improvement would be in the user's understanding of what it can do. Having that information available to customers more easily might be helpful, although I believe it is available to customers easily through their website. There are so many pieces of their product that integrate with one another that perhaps a recommendation for improvement would be some sort of bigger overview and map to help understand how all their pieces integrate together. On the plus side, they always have resources available, instantly, when you have a question about it. It's not like that information can't be found.
I'd like to see the nGeniusONE, the nGeniusPULSE, and the OptiView, their three separate products, work a little better together, a little more streamlined. We can hook up an OptiView to our system and it will bring it up on our nGeniusONE splash page where we can go and click on it. But we can't really use the OptiView functionality with the nGeniusONE functionality as far as throughput tests go. If we wouldn't have to have multiple OptiViews throughout our system, and we could just have one that connects straight back and does all the functionality with nGeniusONE that two OptiViews do, that would be awesome. Another thing that would help out is if they packaged the NetFlow monitor into nGeniusONE. Their NetFlow monitor works with nGeniusONE where you can actually get the netflow of pretty much anything you hook it up to. But it's a separate box that you have to buy. If there was a way that they could package that into nGeniusONE, it would be a complete package straight out of the box. It does a lot for you without it, but with the NetFlow monitor, in our situation, we'd be able to replace three other tools right off the bat.
The GUI has gotten better over time but there could be some improvement in how the GUI is built. That's one of the major areas of feedback I get from the users.
This is a typical thing, but every time they do a major code upgrade, we get hit with some nasty bugs. Some of them literally stop the whole platform from collecting traffic data. They should really do more Q&A on software stability before release.
Most of the functionality I mentioned above could be improved, to be honest. Also, it's not intuitive, it's not simple to use. It is probably the only monitoring tool, out of all the ones that I have, that I really need an expert on, an expert from nGenius, a contractor that I have to pay, to manage the tool. And that's because it's simply not easy to use. NETSCOUT needs to focus on making it easier to use. I should not need to pay an expensive resource to be able to manage the tool for me. With any other tool, I'm able to do that management internally. They should focus on the user experience, not just on the capabilities that they can provide. User experience is important these days. That would be one area where it could be improved. Another, which might be related, is that it's almost like "white elephant." There are so many features that it makes it hard to know it all. You end up paying for things that you don't use and probably don't even need. It might be better if NETSCOUT came up with a modular way to pay for what you're going to use and not pay for all of this "white elephant" without being able to take advantage of all of it. I also think that it's a little too dependent on physical agents all over the place. If they were able to move a bit more to the virtual environment that would be better. I believe that we still depend too much on physical appliances to get the most out of the tool. And by the way, I recently found out that they do have some virtual environments that they can deploy but I'm not sure that it's widely known yet. Those are the main areas that I would improve.