It's a fairly robust, infrastructure-type monitoring. It gives you a framework and lets you implement custom code, which can work well. It has some pretty good documentation, although not all aspects are as well done as I would like. It’s also extensible.
Because it generally meets our needs on the whole, I’m fairly satisfied.
We have a fairly big IT department with different teams being on different calls. We use OpenView to watch databases, apps, Windows servers, and node servers. Even though it’s not monitoring application health directly, it still sends notifications to let teams know what's happening.
It just makes it possible for the on-call tech to relax and wait for stuff to notify them of problems instead of having to, over the weekend, constantly sign in. It babysits for them, when we code it right.
Having used many solutions including SCOM, I agree it works well with Windows systems, but I don’t think that *NIX Administrators or Network Administrators find it as user friendly as other monitoring solutions, aka Nagio, Big Brother, Solar Winds, while it has improved, it still uses custom MSSQL database while others can use many different types of database solutions like Oracle or open source databases like Postgres. With companies increasingly looking down the path of open source products, it is limited in that area.