Is it possible in NPM or SAM, to have a new volumes be automatically monitored? The current process is that if a drive is added to a Windows server, someone has to go into Orion, and check off that new volume to be polled now.
I'm looking for something that will either automatically start monitoring that drive or to alert my team that a new drive was added, so everyone is reminded to add it to monitoring.
Same issue here. Setting up an auto-discovery set last time took 30 jobs. We have very slow links on nearly 600 remote sites and the jobs take forever!!! to run
Even though importing volumes can be automated now, you need to go in and delete the dead volumes.
I did a discovery against 900 nodes last week and it imported about 20 "ghosted" disks.
The problem with these disks is the only way to get rid of them is to delete the node and reimport.
Thus, losing al historical data for some of the largest financial companies in the state.
Ya'll wonder why I sound angry well it's stuff like this that really makes me despise NPM.
Do you realize how much of a PITA this is?
OK, not trying to be dense...
what do you mean by "ghosted disks"? Disks with identical "unique" identifiers so you have duplicate disks.
I don't get what you mean...
If you rename volumes or add a new volume with the same drive letter as the old NPM seems to get confused with these objects, reports show volumes that don't exist in Node Resources.
I have had this happen with two independent environments I manage. Engineers will change and update at the server level but not update Solar. Inconsistencies begin to inevitably happen. Disks don't go away.
The only way out of this is to delete and re-add the node.
Solar should be able to inventory a server on a schedule and articulate the number of VALID fixed disks - our issue at hand.
I can post a screenshot of this. But I am sure other people have seen this.
My problem is the disks seem to stick around in other tables and so there's this weird inconsistency between reports, node resource lists and the main node page.
We had a volume that would not go away from our daily disk report but it did NOT even show in his resource list. We had to delete and re-add it. Now this client loses all availability data. Wonderful right?
This is why we do backups. If we need to build reports on old volumes you can do that using using your backup data. MAaybe not ideal, but no tool is going to understand the difference between a drive that is down vs removed.
This is mainly an issue on Windows VMs. We are going to look at auto-discovery and scripting an API job yo find volumes dead for over x days and delete them.
That is like the old process of custom properties to mute alerts that has now been incorporated into the product.
We'll engineer something that should be core functionality.
Well, 5 years later, I doubt this will get serious traction. Its not just about disks, its about the whole product for large environments is a challenge to maintain without the automation like that which SCOM provides. We tried doing the scheduled discovery but for 1500 servers with 30+ clients and 60 different logins, its a nightmare beyond nightmares.
I would really like this feature as well.
We had an issue recently where an alert was thrown for a database running out of space. In fact it had been moved to a new volume that was not being monitored and it took awhile to figure that out.
I believe this should be part of the normal rediscovery polling (that is used for interfaces, is my understanding) rather than a scheduled discovery... which has the pitfall of forgetting to include new nodes or running into scans so large they time out.
I'm in the same situation. At least, I would like to create an script that when a volume change its ID and in Orion appears as "Lost or Disappeared", it triggers an alert that execute an script with the intelligence to remove the old volume and add the new one (that has a different ID but has the same letter).
Has somebody been working on these??? Thanks in advance.
What's more! I saw that for interfaces you have Orion SDK actions that you can execute to discover and add an interface resource in orion:
- DiscoverInterfacesOnNode (receive NodeID as parameter)
- AddInterfacesOnNode (receive the NodeID and the Interface to Add, taking into account the result of the previous invoked action)
If SW team make available in a new SWIS version with the following Volumes actions, this problem can be easily resolved with a simple powershell script that can be run when a Volume appear as "Lost or Dissapeared":
Shall I create a feature request for this???
Yes ! PLEASE ! The detect device interface is what I dream about because my network admins enable ports on switches but never go back to the node to enable that new interface to be monitored, so we go months without monitoring a new interface until something bad happens.
Can we please get this feature?
I dont want to monitor all interfaces as some are Laptops and when they leave, we get a false indication that the port is down when the port is a laptop but we still want to monitor it when the laptop IS connected.
If this can already be done now, please tell me how to do this. Thanks
Nope. The only thing you can do is maintain a static list of servers and create a discovery, and specifying to only import volumes. Its a terrible process for us.
We have been burned by missed disk alerts so many times due to new or relabeled volumes not being monitored.
I will take this basic "feature" over PerfStack / AppStack any day.
SolarWinds solutions are rooted in our deep connection to our user base in the THWACK® online community. More than 150,000 members are here to solve problems, share technology and best practices, and directly contribute to our product development process.