What is our primary use case?
We are a company who specializes in analytics, both testing and analysis. We get involved in DDI: DNS-DHCP-IPAM. We do a lot of stress testing, protocol testing, and protocol analysis, so we use a number of different tools. We typically try to promote to our clients best of breed type products.
At the moment, we have a demo system with the whole suite. That would be VCX and PVX. We are using the Skylight centres to do link testing as well as analytics for multi-office simulations. So, we have set up a demo system within our lab to talk to a couple of home offices as well as look at traffic coming into our own office. Therefore, we have some setups that we can demo to our clients who want to do similar sorts of things, e.g., people who want to be able to text links or look at jitter and latency from one site to another.
Typically, people who have multi-office/multi-sites want to be able to link test and understand the latencies between each site. First, they want to be able to understand if there are any throughput issues. Second, from an analytics perspective, they look at application network performance from a centralised perspective.
We use it for demonstration purposes to provide solutions to our clients looking for this level of analytics and level of testing. That is our primary goal. While we use it to look at our own networks, it is mainly there as a demonstration type of tool. We have a number of tools that we use for demoes. This is just one of them. However, the fact that we have a demo system to show people is also the way we look at analysing our own network. We can look where things are being held up or being impacted, such as cloud apps, on our own in-house environment as well as at our connection to our public service provider.
We have a public cloud instance for all our other probes or SFPs. Then, everything else is on-premise, VM deployment. For example, even our PVX analyser is sitting as a VM. So, we have one big server that is partitioned out with different VMs to handle all different parts of the Accedian suite.
How has it helped my organization?
We are dealing with a couple of organizations who are using Skylight sensors, so the SFP based analytics to look at multi-sites.
There is a large supermarket chain looking to invest and put into deployment a lot of sensors to look at different store performance, trying to understand latencies between stores, packet loss, and why certain applications are having issues. They are really looking into the link metrics of latency jitter, packet loss, and response times for multiple sites and stores to give them some insight into performance of networks, since one of the things they have got to work out is, "Do their internal networks have a problem? Is it their external networks? If it is their external networks, does it have anything to do with them, their edge devices, or is it the carrier who has the issue?" It is about the finding the demarcation between internal edge and carrier problems to understand who is at fault. This is sort of an example of where our customers get some value out of the solution.
It is about finding operational problems. When sites go down, we try to determine who is at fault. While there is not much finger-pointing, the solution is just trying to analyse when there is an outage and where do we start looking to fix it. The very nature of why organization chooses to use the solution is to accelerate the meantime to resolution and find where problems lie to get them rectified as quickly as possible.
It helps optimize productivity and downtime. Before our clients used this sort of stuff, the downtime was sort of an unknown problem. People were doing silly things with Wireshark traces trying to work stuff out. The hard part was finding out what the problem was. There was meantime trying to understand what the issue was, then trying to solve it. We have cut down that scenario of trying to work out where the problem is, which is obviously a benefit to our clients who use Accedian sensors.
What is most valuable?
We use the whole suite. The Skylights centres are doing a lot of the regular work. Looking at site-to-site latency between home offices and the office, it is sort of running every 10 seconds getting some analytics. That's giving us a really good understanding of our networks and network speeds from our home offices to a central point. We are sort of testing the performance of the service and our service providers at the same time as looking at connectivity between two or three sites.
We are using Skylight PVX to do analytics for application performance and troubleshooting.
A lot of the stuff that we have seen from Accedian, especially where they're heading, they are trying to amalgamate a couple of different products from acquisition and different technologies into one centralized GUI. Even though there is a bit of a difference between both GUIs, we find that the widget type based GUI is really easy to use, as it's informative. You can graph overlay as well as do statistics. It is also really handy with the Skylight PVX analytics and their dashboards. They already have some standard dashboards quite useful for giving you a good 1,000-foot overview type of scenario.
Accedian's deep dive tool set is a slightly different GUI, because you are sort of changing the GUI on the deep dive, but it is very detailed and you can drill right into the heart of the whole conversation, having a look at all sorts of metrics, retransmissions, TCP problems, etc.
The analytics are really good because the analytic engine is good, and they are working in the right direction with it.
They have their security working well. They have a security plugin going into Splunk, which is good.
What needs improvement?
I would like to see some improvements in parts of their synthetic transactions, which includes all the latency, jitter, and throughput. I would like to see some Layer 7 analytics in there. I want to be able to do a DNS request, HTTP GET request, or even SIP call point-to-point or via registration.
They don't have Layer 4 to Layer 7 elements, which I would love to see, because we are being asked to give people so much, e.g. network latency. network throughput, jitter, Layer 3 and 4 type metrics, and stateful traffic (to understand state fully). For example, we have seen jitter and packet loss. How is that actually affecting a real application, TCP, PCP windowing, resizing, and are we going back into a slow start? I'd really like to see that sort of stuff. We've done a lot of testing in the environment of Layer 4 to Layer 7 from a stress testing perspective. We would love to see that sort of synthetic transaction being brought into the Accedian toolset because that would probably give them a more complete picture of everything network and application if we could do synthetic transactions.
For how long have I used the solution?
We started using it at the beginning of the year, just as COVID-19 started to crash down on us.
What do I think about the stability of the solution?
Stability seems to be fine. We haven't had any issues with our staff or demo staff. We probably tend to fiddle more than other people. The thing about using demo labs is that you tend to sort of put those environments under a bit of stress, because you're changing, deleting, and reorganizing things sometimes for different sorts of views. However, from a reliability and stability side of things, you don't have to sit there and reboot your browser, refresh your screen, or reboot boxes. It has been pretty much turn it on and keep on going.
What do I think about the scalability of the solution?
We haven't come anywhere near the performance limits of the tool. You can put multiple VMs to handle multiple flows, so it's ultimately scalable. Skylight sensors with SFPs can be rolled out with as many as you want. It comes down to what your license is.
How are customer service and technical support?
I haven't had to use the technical support that much. When I've gone in and spoken to support, it's been really quick. Locally, there are SE guys here that can look after us. Everything we ask for, we get answered quickly and efficiently. There are no dramas from a support perspective.
Which solution did I use previously and why did I switch?
Their UI is actually similar in some ways to what we used to use in a product called TruView from Fluke Networks. That was a really good product that unfortunately, when NETSCOUT bought them, they killed off. They really dropped the ball on that.
We came from using Fluke Networks and some other tools. We were involved in NETSCOUT, but moved away from them because the solution was troublesome and problematic to scale out easily and build for customers, who were getting frustrated with it. So, we found Accedian tends to be quite easy to use. The troubleshooting aspect of it is all there. With the troubleshooting, you don't have to go and configure from scratch the troubleshooting views and easy ways of looking at troubleshooting, as you do with some other tools, especially the NETSCOUT type of tools, where you have to build all the dashboards yourself. The standard dashboards and troubleshooting information are really easy to use.
How was the initial setup?
It is quite straightforward and very simple to set up.
We were having some issues setting up a couple of centres because we have NAT firewalls. We were having some issues getting connectivity to some of them. Therefore, we were using Accedian to look at other problems in Accedian to troubleshoot, e.g., what ports are being requested to be opened, then work out some of the firewall NAT rules. So, we actually used Accedian to check out Accedian. It was extremely easy to use for troubleshooting.
If you want to start looking at different types of metrics or adding a site, you just have to program it up in SFP, send it out to site, get someone to plug it in, and redo it. So, if you want to start putting more sites in, you could probably do that relatively quickly. If you want to start looking from an analytics perspective of doing more detailed analysis, setting up IP ranges for different sites and types of applications, then that would take a little bit of time.
Generally, even in dashboard creation and certain dashboard configurations, you might want to do a little tweaking. You could probably spend another two or three days just doing the finer points, but the beauty about it is all the data is there and been collected. You can basically push in the new definitions and get that information back into it anyway. You are not losing data, as it's all been collected. It's just a different way of viewing it. After that, once you've got it going, it's set and forget unless you want to change to a different type of application, roll out a new application, or have a different site coming online.
What about the implementation team?
We did the installation of four VMs in a cloud environment. Accedian set up the cloud environment. They give you the credentials, then it's just a username and password. Once you're in, then you're in, and it's just a matter of setting up all the VMs, which either com as VMware or KVM to download and install. You pretty much open up your virtual machine, import your VM file, and away you go. It's pretty easy.
Usernames and passwords are all defaulted so it is just a matter of getting in and setting IP addresses up, whether you want to be DHCP or static. Once they are up, you apply a license and you're running. We set up four or five VMs in the morning once the cloud account was set up, then we pretty much had traffic running through it within a day.
To get our whole demo lab up and running, it took me a day. I set it up myself. I got the VMs provided to me by the team that provides the download links.
The download links were quick. We were all running on 100 Mbps links, so I downloaded all those images pretty quickly, then installed them. The slowest thing was probably going through and just setting up all the IP addresses because I run everything statically assigned.
For a couple of hours, I was working with one of the guys from the city just going through some configs and getting some configs setup. I was pretty much a fairly green user of the product at that stage, though I've used a lot of tools like it.
What was our ROI?
At the end of the day, it all comes down to solving network issues and problems quickly. So, it comes down to engineering time and return on the investment.
What's my experience with pricing, setup cost, and licensing?
The pricing of Accedian Skylight is really good. The sensors are low cost. Their model to analytics for sensors is by license, endpoint, or session. With the probes for their analytics, if they get deployed virtually, they are free. The licensing is only based on flows. So, you can effectively deploy probes everywhere in your network. Then, if you want to look at a specific type of traffic, you can enter into it with a very low cost license. You can just use things like spam ports, mirrors, TAPs, and aggregators to optimize what sort of traffic you send to these analysis tools. Then, if you want to start looking at more, you can up your licensed as you go. You are not getting forced into expensive appliances or subscription models.
It is a really good pricing model. That is one of the things that we really like about offering it to out clients is that they can start small and grow as they need, but they're not left out getting the right sort of information straightaway.
Which other solutions did I evaluate?
We use a company called Colasoft, which is a bit more focused on network analytics, capturing all packets, and doing detailed packet analysis.
We have a relationship with ExtraHop, but that's probably more from a site security standpoint than it is from raw analytics. ExtraHop does sort of compete a little bit with Accedian, across more from an analytics side of things. We have used it more from a security standpoint on getting some visibility with a lot of security things. However, there is not a lot of overlap with Accedian.
We regularly work with a company called Profitap, who does a lot of network tapping, packet brokers, etc. We use products like Profitap for front-end traffic collection to feed our analytics tools. They sort of compliment Accedian. They are a TAP aggregator, which feeds data where you can do filtering and data manipulation along with killer traffic brokering or packet brokering.
We constantly look at other vendors all the time. Other vendors come to us all the time wanting us to look at their toolsets for partnering and things like that. We've looked at a few, like Gigamon, but we are sort of happy with where we are at the moment.
We don't play with SolarWinds. I think SolarWinds answers SNMP type issues for people. Their approach is getting very cheap, e.g., it starts to escalate from a price perspective, then before you know it, you spent a lot of money on SolarWinds. Therefore, we don't get involved in SolarWinds. It is not the right fit for what we do.
What other advice do I have?
We are a relatively new partner. We certainly haven't explored the full gamut of the suite. There is a lot more work to do, especially with dashboard creation, dashboard optimization, and customization. In terms of the sensors, we're only using that in a very small manner. We are probably 25 to 30 percent along with our journey in that scenario.
We do use the solution for performance and traffic monitoring of cloud environment as part of our demo set. We are able to look at cloud apps and cloud environments, more from a monitoring solution. We don't actually have anything virtually in the cloud at the moment. We have been using response times out of things like Office 365 and some of our CRM tools, but we haven't deployed anything in AWS or Azure. We can look at the traffic, where the response times and those applications are being hosted, etc., so we can sort of break down what applications are performing well.
We provide tools to our clients based on answering issues that they need. We see a differentiator between this toolsets and others. It gives us some insight into helping clients size up network and analytics solutions.
To get value out of a tool set, you need to have an idea of what you are trying to achieve to start off with. We see so many people who just want to stop monitoring traffic because there's no objective to monitoring traffic, and there's no objective to what they're trying to do. They just go, "We just want to see what's going on." Half the challenge that we have with people is helping guide them along the way to work out what they want to see, how they want to see it, how they want to presented it, and what sort of information are they trying to looking for:
- Are they interested in application issues?
- Are they interested in network issues?
- Are they interested in trying to work out application versus network?
- Where do we start looking at things?
- Are they trying to look at it from a set of sites?
- Are they trying to look at it from the users' perspectives?
We try to understand what they are trying to use it for. There are toolsets out there that will do a lot of things for you really well. Ultimately, you need to work out what you're trying to achieve. If you're just trying to look for up-down scenarios, then Accedian will not give you too much information. However, if you're trying to understand where the latency in your network is, where you're seeing packet loss, and when you're seeing bad response times, then that's where these sorts of tools help you. One of our things is we just try and help people understand what they are looking for and guide companies about what to do. At the end of the day, we're just trying to help people make the right decision so everyone's happy in the long run.
They are a solid eight out of 10. If you look at the way that they do link-to-link testing, Accedian does it better than anyone else. I also like their Skylight sensor, TWAMP testing, and analytics. They do need a little more work on their analytics and adding Layer 7.
Which deployment model are you using for this solution?
Which version of this solution are you currently using?
VCX and PVX