mesverrum · Observability Architect · ✭✭✭✭✭

it would only affect things in NPM that were relying on syslogs/traps. I find that 90% of my clients shoot a mountain of syslogs at orion and never look at them again, but you'd need to know your environment to be sure.

Haha, I scoff at your "cannot be done" If solarwinds has the data in it you can present it for sure, sometimes it takes some digging around but it's always doable. I'll try to find some time next time I'm in an environment with UDT running to see what I can spin together. What Info do you need in this summary? -Marc…

You can ignore that first paragraph in the KB, thats just letting you know you have the option to make the monitor retry if it fails. I've never had a use case where I needed to edit that file. Just go to the manage templates screen in SAM and assign the built in SSL Certificate Expiration Date template to any server where…

For whatever reason the only threshold verbs we have in the API right now are to set the thresholds for hardware health sensors. Hopefully that gets cleared up in later releases.

Thats going to be horizontal though, if you want to make it vertical it is a bit more complicated, but the trick is to build unions like so: SELECT 'Comment' as PropertyName, i.CustomProperties.Comments as PropertyValue FROM Orion.NPM.Interfaces i where i.interfaceid = ${interfaceid} union all (SELECT 'CarrierName' as…

If you have defined a dependency relationship for each site then when the parent goes down Orion stops polling the children and sets them to Unreachable status. As long as you don't have your node down alert logic including Unreachable status objects then that should cause the children to stop generating alerts. There are…

And as you said, SNMP support for Meraki is very bad and their API is very limited (although got a bit better recently) and that is something only they can fix on their end. I found recently that you can now query the Meraki API to get the direct link to the dashboard page for whatever object you are dealing with, so I…

I'm not seeing the use case for generating a new file every minute for anything that is polled from orion, especially since the vast majority of polled data is being collected on intervals between 2-15 minutes. Why wouldn't you just set the polling intervals to collect your metrics every minute and then run a report…

No need to resort to powershell and scheduled tasks, The easiest way to do this would be a scheduled report of discovered nodes, but if you want to get really fancy with an alert so it only notifies you when something new is found you could do it like this. Set up a new alert with a custom Swql alert condition like below…

It's not corruption, it's normal. The alert preview does a select distinct type thing and only shows you unique names. It is super obvious when looking at application alerts because most instances of an application template will have the exact same name so the preview says only 1 will trigger, in reality you get alerted by…

I think on that particular occasion we just called in while we were getting ready to begin and waited until we got someone on the line.

I usually prefer to do queries in a message body since converting the strings for http can get messy at times https://loop1.com/swblog/using-the-rest-api-to-get-the-most-out-of-solarwinds-part-1/

I do like this work around since critical is not normally used for node statuses it wouldn't be as confused with another condition. Still kind of a clunky process until the atlas application itself supports grabbing the value you need. In any case I think it's super important that custom sql/swql queries be supported in…

My method to do that is just to copy paste the objects that i want to match, then right click its properties and change the object associated with it. When I'm doing a LOT of maps i tend to just get all my preferred graphics set up how i like them with the appropriate variables in their captions and such and I save that to…

You can just add custom query as a resource to the Web report, most people think you can only do table or chart but any orion resource can be used on the reports if you look at the box on the left where it asks you if you want a table or chart

Expanding on the value of writing events to the log, I have gotten to a point where for almost any custom reporting I do I no longer actually write the report to scan the actual tables that the metrics live on because in large environments I'm scanning hundreds of millions of rows and this can lead to reports timing out…

Add a firewall rule to your db server blocking that ape for 15 min and see how much data ends up in the msmq storage folder and project that out? If you have NTA and a layer 3 device in the network path you could try to measure the conversations between the ape and database server for an hour. Run wireshark or another…

Yeah I ran into the same when I tried it, the detailsurl works fine for the details but not the LED. I'm sure there's some specific syntax it is looking for to let it know it can display that status with the status icon watermark but I couldn't find it. You can force the issue by fully writing out the whole html tag and…

I don't typically use python so I don't see any obvious syntax errors, the remove operation only requires the URI, so I'd probably print out results['results'][0]['Uri'] to double check that it's being output correctly. 

Normal as in a mistake that the tier 1 support frequently get calls on and have given up on fixing maybe? But obvious not the intended function of the tool and also not "typical" in that currently none of the 5 environments I can access exhibit the issue... Sorry to hear the lack of progress.

I typically handle the displaying disks separately because they can make a real mess of the table but while i was working with my client today I did end up coming up with what I think is a relatively clean solution to the disk volume status problem. Normally if you join the volumes table you end up with a row for each…

So select is picking columns that we want to display, the tables we want to use are defined in the from and joins. From is your starting point, then you join other tables that have related information by whatever values they have in common. You can rename tables to save yourself from having to type the full names out, so…

You can see where it went wrong here You are looking at the current stats on the interface table, needs to be interface traffic history

I've deployed full acm/nam environments, from the installation and hardware requirements side it is not really any different than deploying the traditional full suite of Orion, there are no incompatibilities our special gotchas. This is essentially just an alternative licensing scheme for environments that would normally…

Agents are tied to the individual server where they live, so you can't really monitor the VIP via an agent. You could probably cook up some kind of powershell script or something to check for the member being the active host or however you would need to do for your cluster.

The retention history for most event data is only 30 days out of the box, so that is why you only see 3 days back. Did you try the version of the query I pasted in? It seems like it should do the job. The datalength function doesn't exist in SWQL, just length,and since all the data you want is on the events you really…

this thread gets into detail on all the wmi port details Portocalypse - WMI Demystified

Sounds like you didn't discover any fixed disks on the nodes in that suplot, need to check that out and maybe redo your discovery.

I haven't poked around to determine how exactly they order the undp pollers on that resource, but I have seen what you are talking about where it scatters the groups. If the example ones you are interested are the same as what you would want in SmartUPS then I would just use the undp tool and edit them and change their…

Sounds like something is preventing you from remote executing powershell on that other system. Does that system have the WinRM listener set up and is the polling engine a trustedhost? Unable to get a remote Powershell script running