Open for Voting

Report to find unmonitored disks

We have a fairly large environment and would like a report to scan all monitored nodes and list any hard drives that aren't currently monitored. We had a problem where someone added a hard drive to a server and forgot to add in solarwinds. When the disk ran out of space we weren't notified. Currently we could either manually check available resources for every server (impossible) or schedule a resource discovery for 40+ IP subnets with over 10,000 IPs and add all detected hard drives regardless if they currently exist (duplicated efforts and time consuming). It would be much easier to have a scheduled report which scans our existing nodes and determines if any hard drives are available to be monitored.

  • Bumping another upvote on this one after an unhelpful Support engagement.  I realize that this would be much less of an issue if I was using WMI, but (as username would imply), I inherited a ~10k node Orion environment that skews heavily SNMP.  I'm changing that one new node at a time, but I keep getting bit by volumes that were added after the fact, or the were being monitored by got extended!

  • You can create a Powershell script to run an SNMP query against the server OID, or poll WMI. Then compare the returned results to the drives monitored. You can even add them via the script if you decide to do it.

  • Yep, that's true. You will need to specifically target this app to a new server. I know SolarWinds team is working on ability to assign apps dynamically, based on rules and policies - this is yet to come as a feature, soon.

    As for you specific example - I have an app that is checking all automatic services are running. So, when you add new services - they will be picked up automatically with this template. Check it out here: Check All Services Which Set To Automatic Mode Are Running

    ... as I said - this is just one specific example, but similar approach is being applied for disks as well (see my comments above).

  • Correct. You cannot target classes like you did in SCOM. So if you create a monitor that targets all SQL servers l, future servers added to monitoring will never know about the monitor. You must manage all of it statically. People don't understand this kind of thing. It's like I'm the Jew trying to convince Muslims to convert. Lol. Never gonna happen. The lack of a class framework is something people just don't comprehend. Zero automation.

  • So it's the same for all resources i.e. what if i create an application template to monitor say, a windows service. Do i have to specifically target that monitor at servers I know have the service? Is there now way to discover this?