Overall Satisfaction with SolarWinds Network Performance Monitor
We utilize NPM to monitor our global network and server infrastructure. It provides us with basic monitoring and alerting capability, as well as historical trends. Until recently, it was used exclusively for our network environment (and sparingly) because we did not have the license capacity to do more. Once we upgraded the licensing, we were able to leverage it for our server environment, monitoring Windows and Linux VMs, as well as our vSphere hosts.
- Generally speaking, NPM does alerting very well. It has a lot of flexibility in how to build and deliver the alerts.
- Being able to do HTTP POST on alerts helps integrate it with modern tools like Slack and OpsGenie, even if it's not a baked-in integration.
- The ability to get all relevant device data from one page (and a couple of tabs) is nice, and there is customization available within that view.
- What I deem as some of the better features in NPM come from integration with other paid SolarWinds tools, such as NTA, NCM, etc. It can be a bit disappointing to look at the marketing material, find something you really want to do, only to discover it requires buying yet another product.
- SolarWinds licensing isn't great. Some of the caveats around "you need x number of interfaces licensed for NPM, so you need that many for this other product" make the suite less and less attractive and the total cost continues to climb.
- There's essentially 3 generations of User Interface built into the current product. Some tools, like syslog, still exist on the old Windows-application-run-from-the-server model and are in the web interface, most things are in the web interface that's been around for a while, and now they are integrating new features only in the newer-style web UI. There's no consistency to those UIs, so it looks and feels like 3 different products.
- The Universal Device Poller is a great way to get functionality out of NPM for products without baked-in support. But there are still some things that seem like table stakes that require the Universal Device Poller - monitoring APC UPS devices for example. Other products (including open source) do this really well since just about everybody needs to monitor stuff like that. Years later, NPM still relies on the pollers.
- Sorry, but I hate the graphs. I prefer rrdtool-style graphs, and there is some hope for the future in the new style web UI.
- We really did not get a lot of ROI from the product for the first few years, but that was our own fault for not 1) licensing properly, 2) getting training, and 3) generally not looking at the data it was generating.
- The amount of noise generated from the alerts caused us to ignore and then scale back the alerts. The usefulness of a monitoring (and alerting) tool diminishes greatly when you stop trusting the alerts. It has become more of a tool for "what happened?" as opposed to "something is happening and we need to look at it". We already have free/inexpensive tools that do that, so NPM is just another tool on the shelf, not the "go-to" tool that it should be at that price.
There are tradeoffs in all of these tools. I really like Observium, but it's alerting seems far below SolarWinds NPM. The Nagios-based tools have tons of flexibility, but require a lot more configuration to provide value.
I didn't select NPM as it was already here, but I've kept it in place as it does generally do what we need it to do.
I didn't select NPM as it was already here, but I've kept it in place as it does generally do what we need it to do.
SolarWinds NPM Feature Ratings
Using SolarWinds Network Performance Monitor
2 - IT - Networking
2 - NPM requires basic Windows Server administration skills, as it runs on Windows, and some level of knowledge on how to monitor the platforms you wish to monitor. If you don't know what SNMP is, chances are you haven't configured it, which would make using SNMP to monitor a device from NPM rather difficult. It does give you suggestions such as "hey this is a Windows box, you should use WMI".
- There isn't anything unexpected or innovative about NPM, or how we use it. I have done everything NPM can do with other tools.
- Assuming we're willing to fork over more money for more NPM modules, there is quite a bit more we can monitor around virtualization, storage and applications.
Using SolarWinds NPM
Pros | Cons |
---|---|
Technical support not required Well integrated Convenient Feel confident using Familiar | None |
- The network discovery works pretty well. It 's definitely better than some vendors I've used
- Even without the virtualization module, it manages to connect the VMware environment with the underlying VMs.
- Some of the connections between the optional modules are less than efficient. e.g. It doesn't assume that network devices should be integrated with Network Configuration Manager, you have to manually select that option.
- The alerting module leaves a lot to be desired. While there are default alerting rules, they aren't necessarily all useful. If the options for email aren't set up properly before you start tweaking alerts, it can become a major mess when you want to change something like the email server address.
- Among all the pretty charts and graphs, there really isn't any indication of *why* something is a problem, or what you should do about it. Great, it's red. Tell me why it's red and what might happen if I don't fix it!