Recent Discussions
Filter within a widget
Hi, I have a custom datasource for some specific SQL metrics. Presently the collection script is explicitly excluding system databases (master, model, tempdb etc.) with a WHERE clause. I'd like to re-include these system databases but have them on their own widget (with the current widget excluding them). Only way I can think of is a duplicate datasource that simply reverses the WHERE clause, but this seems way overkill. Can we filter per-widget?
collector installtion on Domain controllers
Hi We have 100+ domain controllers where we want to install local collector to monitor them.But everytime we need to execute the command on server one at a time,is there any atuomation possible which can be done to install it on all the Domain controllers at once ,as the Collector ID and the token will always change and we cannot use the same installerDoes a change in a Dynamic Threshold reset the training data?
We rarely use dynamic alerts but someone wanted to test for it. We set something up but accidentally left the polling interval at 5 (50 minutes). It happened to be set that way for a few weeks before we got around to testing an actual event that exceeded the band we set. But it DID work, it just took almost an hour to get the alert. Well today when we realized this, I tweaked the dynamic threshold, mainly just changing it to 1 interval (2 minutes) because we want the shortest possible response. But the engineer also dropped the bandwidth on the port to a super low value at basically the same time. Its been an hour or more, no alerts. But literally all i changed was the qualification to make it happen sooner. So the TL;DR is ..... does a change to a threshold make the training of the datapoint start over? I see references to how long it takes to train, but nothing on whether changes reset the training, or anything that would explain this.
Making "Services" usable/scalable
I've been bumping into tasks lately that would be great candidates for using an LM service to combine data into a more usable set pulling from other DSs without having to duplicate effort. Adding a ServiceSource in the modules would solve for it... the ability to add the 'Source to a group instead of a device could allow it to show up as a device in that group with the service icon. The primary issue I'm having with the current implementation is that it's a single thing that doesn't scale to "all client environments." As an MSP for a specific industry vertical, many of our clients use same/similar software stacks. Making a service means I repeat the same thing 50 more times to make it for all of those clients. I'd much rather build once, then be able to apply it to a client environment: /Clients/* and have it propogate itself, discovering the appropriate resources via some kind of active discovery type of mechanism. Just a starting point for conversation at this point. This would allow for implementation of application monitoring for third party application multi-tiered enterprise stacks where we don't have access to the source code to add monitoring hooks.
NetScan via CSV import Description=xxxx
So while doing a netscan CSV import it is possible to populate the "Description" field on each device someway? There's only 4 columns but but nothing for Description details. I have over 300 devices that I need to add the description to and would be great if one of the columns was for Description. I see no way to do this but just wanted to ask.MS DHCP DataSouces need Explicit Sessions
I've posted about it in the past, but using remote powershell in a windows environment with an implicit session causes a profile to be loaded even if you tell it not to do so. These all need to be done using "-session" rather than "-computername" to target the remote entity. This will allow setting the no profile flag in a way that actually has the expected behavior. This has caused issues in the past overloading the mount request queue and breaking profile loads for normal users causing them to receive temp profiles rather than their own. In this case, luckily, it doesn't tend to target systems that users would be logging into... but this is the mechanism that allowed me to help uncover a thirty year old buffer overflow bug with windows. (MS didn't give me a bug bounty :( ). I'll look up the thread I had here and link it in a follow on comment.UIv4 Module Code Editor/IDE funkiness
code reverting (reproducable): If I make changes to some powershell in the new code editor, then change tabs/windows. When I change back to the window with my editor, it has reverted to the previous save point. An Undo will bring back my changes... but it's scary after making a few changes after staring at your code for longer than expected working through logic in your head. property source error output: In UIv3, there was side by side output vs. error screens. We now just get the little info bar at the bottom that states "couldn't do"... but not where/why/what that could be troubleshot. I'm having to add in a ton of troubleshooting output to the property definition to be able to expose the data I need to fix the scripts. The one I just completed is nearly 30% error output handling to be able to figure out what is wrong with it. Other than those two things, I'm pretty happy with the changes from the old.
