What is our primary use case?
Our installation is on-premises on a virtual machine. When we purchased NPM I was particularly interested in the features of Orion. We were going through a maturation process propagating the monitoring of applications throughout the server policy here at the university. It was more to monitor various applications that are primarily used for individual schools. For example, we have the school public health, and we have the school of nursing and they have different needs. You have all these applications like Banner and Canvas and other web-based applications that get used by particular areas. We are working with one application now for COVID-19. There is an application that has been spawned from that, but mostly our monitoring is for up/downtime, space, memory, availability, and those types of things critical for performance analysis. We are just now getting more into monitoring databases and a broader scope of applications as well.
What I am mostly referring to is not the whole Network Performance Monitor suite. This is the actual operations monitoring piece of it. We have the network and we are responsible for set up and monitoring of those particular tools.
What is most valuable?
We identified the five areas that we wanted to have be our key performers, like memory utilization, CPU utilization, and those core components. Just the basic care and feeding of servers, whether they were physical or virtual.
What needs improvement?
We were looking at using DPA (Database Performance Analyzer) which is another component of the NPM and that looks like it takes care of an area we had not really dove into yet. But in talking to the DBA, they like another product that they want to use for the database monitoring. It is an Oracle product called OEM (Oracle Enterprise Manager).
In talking to the DBA's, there was a lot of functionality that they were just used to and they had issues because they did not see some things that they thought were important but missing in the DPA product. That is the only thing I have run into so far with the product that seems to be a shortfall. Everything else is pretty good.
I do not really have other suggestions for additional features beyond working on the development of DPA to make it a more complete tool. We are still learning the product.
There may be some things that they could improve in the reporting aspect of the product. You can create custom reports and those types of things. But the product seems to be missing some depth in that area and it leaves a little bit to be desired. For example, the filtering system and formatting reports and things like that seem to have room for improvement. You are a little limited on the information available on that front and the utility could be improved.
They do have some templates that you can use for report setup. It would be nice if you could maybe make your own. That might make it a little bit more easy to generate other reports that follow the way you like to do things.
For how long have I used the solution?
We purchased SolarWinds NPM (Network Performance Monitor) about three years ago.
What do I think about the stability of the solution?
SolarWinds is a pretty good, stable product, but we have had some issues with support. That is especially true of the last release we upgraded to. There were some SQL issues that we were looking for them to help us out with and, finally, it took so long that someone on our internal SQL team had figured out what the problem was. So, I would say that the upgrades to the product itself from what we went through can be somewhat of a pain point and maybe there needs to be more rigorous testing.
What do I think about the scalability of the solution?
Typically, I don't think scalability is an issue. If you are talking about individual users, we have anywhere from 20 to 25 people across the typology. Probably one of the biggest users is a resource/research support group where they have various servers on the Orion Monitoring System. They actually take on their own responsibility and monitoring role for their servers. So, there are some groups that adopted monitoring through Orion, and then there are some other groups that have quite a bit more of a blasé approach to it and do not really think it is important.
How are customer service and technical support?
We have not had a lot of opportunities to use technical support. When we did I was not particularly impressed with their response time or expertise.
How was the initial setup?
The initial setup with Orion was somewhat complex. They do have templates in there that you can utilize, but most of those you are going to have to customize anyways. We found that it was easier just to go ahead and do our own.
The way that I found it to be complex was that we had to do a lot of our own scripting to collect the data that we needed. Once you have got a template set up, you can pretty much populate that information to the Orion database, but I mean there is some pain to get started off. Overall we got involved in a year-long project where we were meeting on a weekly basis discussing all the issues. I think it took us a little over a year-and-a-half to actually declare success on the project.
Only after that declaration of success were we able to get on to monitoring and dissemination of information to various users and get started with the monitoring. But our monitoring is a lot better than what it used to be. We were using Homegrown, Nagios, and Check_MK. People were not paying attention to alerts. I think the way we handle it now has really helped users mature in that way. We have still got a lot more work to do though.
We started out with an implementation strategy, but there were some challenges when our directions got changed on us by senior leadership. We would have a plan and a direction and head down a path to implement those particular types of solutions. We would set up the alerting and monitoring and then three months later somebody would come up from an executive level and say, "Hey, we need to monitor this as well." So, we would have to make some changes just like any project. We had to learn to be flexible with it to meet people's needs, but having clearer goals from the outset would have made it easier and quicker in the long run.
What about the implementation team?
We did just about all the work on the implementation on our own with an internal team. Right now, I have got one person that is in charge of the actual product itself. They take care of all the changes, releases, upgrades, deployments, the agents, and those types of things. I have another individual who is responsible for creating tickets and disseminating the information to the owners of the server monitoring in various groups to get their alerts fixed. It does require attention but it does not take a large group for maintenance requirements.
What's my experience with pricing, setup cost, and licensing?
I do not handle the negotiations on pricing, so that is outside my realm of influence. I know that our initial purchase was about $200,000.
What other advice do I have?
One of the things I would suggest to people considering this product and products like it is having a clear agreement upon targets of what you really want to monitor. If it is applications or servers or whatever it is, it just makes sense to come at your choices with a plan. Implement that on the front end by picking out what is important instead of coming back on the back end and trying to do things after-the-fact.
On a scale from one to ten where one is the worst and ten is the best, I would rate this solution as about an eight-out-of-ten.
Which deployment model are you using for this solution?