How has it helped my organization?
We've had a very quick return on our investment as far as being able to monitor applications that we hadn't been able to monitor before, and then able to catch problems that we previously had not been able to catch. For example, with the way we have AppPulse deployed, we were able to catch a problem at a particular call center where the application was running roughly 10 times slower than other call center sites. We were able to catch that and help remediate that. It affected a lot of call takers with very slow-running application and we were able to help resolve that quickly.
What is most valuable?
AppPulse Active allows us to very quickly produce synthetic monitoring for our applications, for some applications we've never been able to monitor before, some older Legacy applications. It gave us a set of tools that we could use to create the transactions and to play them back flawlessly on applications that we previous hadn't been able to. It was very quick.
It can monitor multiple types of applications. We monitor an old client server app, our mobile application, our external-facing web application, and internal sites that we use, like Skype for Business. Very quick and easy, and covers a big breadth of applications that we can monitor. It's very important to us because we're a regulated utility and we're required to have certain response times to calls from our customers.
One of the applications we're monitoring through AppPulse is our call center application. We want to make sure it's performing well and available so those numbers are good and we're in compliance with the regulations from our governing bodies, from the commissions. That's extremely important to us. We need to know our applications are performing and performing well.
For our external-facing websites, it's not regulatory-bound, but our customers are getting to demand better performance as they use other people's websites, they buy stuff from big sites such as from Amazon. They expect our site to perform equally as well. Also, if our site performs well, some basic questions could be answered online saving on calls. We need to be able to do that on our website to keep our call center volume down. It's all about keeping the customers happy and making sure that we're in compliance with our regulatory bodies.
What needs improvement?
The one that sticks out in my mind is some flexibility to do recurring downtime. Right now you can schedule downtime on a one-off basis. I'd like to be able to schedule application downtime if I know the application I'm monitoring is going to be down, and the one I'm responsible for needs to go down every night for a scheduled recycle. I have to schedule it everyday manually, but I would love to say, "Every night at 2:00 in the morning it's going to be down, so don't monitor it for an hour." This would prevent false positives saying that the application's down. That's one thing I would really like.
Also, I would like more granular reporting capability, and a larger breadth of a time period that I can run a report for. I'd like to get more granular with the data points that I'm capturing as far as what my performance metrics were. Right now it's averaged, so I get an average. I think I get some standard deviation from the average. I would like the actual data points if I need them. I would like to have that data broken out by the machines that I'm monitoring and not rolled up into one data point. This would enable me to see trends with actual data as opposed to just averages.
We monitor applications, for example, from five different machines. I'm running the same transaction on five different machines, but when I run the report it's aggregated, but I want to be able to split that up. That would help me know how one site is performing over another. I can get it at a high level but then I like to be able to use the granular data to maybe back into when exactly things start going awry, which I might not get from just an average and standard deviation. Having these data points maybe helpful to our development team with some analytics, from a transaction standpoint, maybe starting to know when a transaction trended towards bad.
What do I think about the stability of the solution?
It's a cloud-based application, so I can only remember one time where it just was unavailable. The actual AppPulse application was unavailable once, and it was for a short period of time. The other bugginess could occur in the external points of presence that they provide us. They give us a list of 50 or so locations where we can run a transaction from. We use those for out external-facing websites. There in different cities on different networks, AT&T and Verizon primarily. Sometimes those points of presence will become unavailable. For example, we use Chicago AT&T and Chicago Verizon because we're in the midwest, those are midwest sites, and two pretty big carriers in our area. We'll see occasionally one of those sites for one of those carriers just stop running the transaction, stop working and become unavailable. That happens occasionally.
We were able to say, "Wow, this site is so much slower than all the other four call centers that we're monitoring. Something's going on." Then we were able to actually use the sniffer tool at our machine at that site to figure out what the problem was.
What do I think about the scalability of the solution?
To us, scalability is irrelevant as it's cloud-based. We're not hosting it, so it makes it easy for us. Performance-wise, I see no issues on performance. The way we're licensed is we buy a certain number of transactions, so if we need to scale up we just buy more transaction licenses.
Which solution did I use previously and why did I switch?
Prior to me being part of the team, I know that IBM tried to create synthetic transactions for our call center application, and they weren't successful at it. What HPE provided was a toolset that works across multiple platforms. We found a tool that worked for our particular platform, we were able to create the scripts to perform the synthetic transactions that other vendors weren't able to do. They were trying to customize something for a particular application, to be able to record a transaction as it was being made, and they couldn't do it and have it played back successfully. Before that, they tried to use WinRunner. Then the attempt with HPE, we were able to do it, and do it quickly.
How was the initial setup?
Deployment's easy. The way it works is that we create a script using one of the tools that they provide, zip it up into a file, and then we go to the AppPulse application, and we create a new application, and upload the file, and it's there. We tell it where we want it to run from. It's very easy. It's very easy to download the script so we can make changes to it. It's a URL we log onto and just create our user ID. We can have a site monitored within minutes.
What's my experience with pricing, setup cost, and licensing?
From my understanding it's relatively cheap. I didn't go through the negotiations of what it cost, but it was a no-brainer from a cost perspective compared to the other tools that we buy as an IT organization.
What other advice do I have?
The first thing I would tell you is to do a trial. You can do a free trial, HPE will give them a temporary license and you'll be able to monitor an application for free within minutes. Especially with web applications. It's very easy. Just give it a try and see how easy it really is and see the data that you're going to get back and how quickly you're going to get it. The only surprise I think you're going to find is that they're going to be surprised at what they uncover about your apps.
There's really not, in my opinion, that much of a downside to it. However, false positives is definitely something you will need to watch out for. Before we automatically send an alert to the operations center, we watched it for months to make sure that we weren't setting the threshold too low. We were able to set thresholds for performance and availability right within AppPulse. If we set the threshold too low, you could generate a false positive because saying that apps not as well performing as it should be when you don't have enough data to understand how the performance is really supposed to be. I think you really need to know your application to set those thresholds.