mesverrum · Observability Architect · ✭✭✭✭✭

correct

This is an example of how I used to add logging into powershell scripts which makes them a lot easier to debug and track what they may have been doing when they failed. github.com/.../DiscoverFilterImport.ps1

You'd need to implement this as a custom script monitor and the script would need to be written to parse the XML

I haven't updated these scripts in a long time, but i have done what you are asking before. This example script has a basic example for muting and unmanaging nodes in the SW API https://github.com/Mesverrum/MyPublicWork/blob/master/SupressAlerts.ps1 and this one is an example of connecting my SNOW CMDB API to Orion to set…

I've built scripts that automatically create their own child objects like you need to do here using the API. Not to distract from your specific question, but why even bother to build hundreds of script monitors to check process state? Thats a built in capability and running the scripts is just a LOT heavier load to do it…

The configs are not stored encrypted and there is no built in functionality to make them encrypted, you should be limiting access to those drives to only people who would already have admin permission in Orion so they effectively already have full access to your hardware. Further, those text files are not actually the…

The netflow tables have always been problematic to write custom queries against. They are extremely common to timeout with people who have any significant flow volume coming in. You might find a good way, but my first thought is that packets might not have an index, so your order by might be problematic. As I think about…

There is a difference in SW between something being red in the GUI vs something triggering an alert, you have basically infinite control over the alerts but only limited control over the way the UI displays things. As soon as a SAM process monitor gets a down response then it is going to turn solid red/down and you cannot…

The way this is done in SW is to create a nested set of Groups, and within the group you can manipulate the status between worst/mixed/best to control for things like redundancy and load balancing. Depending on how specific your logic is these nested groups can get pretty complex, but as an example using SW HA and my SQL…

 @"cscoengineer" Has done pie charts in the custom html panel with javascript several times before, of course today I can't find a link though. Did find @"adam.beedell" posted some working javascript in this thread a few years back thwack.solarwinds.com/.../59302

So i think there are a few ways to look at this. Is the SCP certification a worthwhile investment? I had lots of experience with SolarWinds before they started certifying people, so the level of effort to study to pass the tests was always trivial. I spend an hour or so the day before the test refreshing myself against the…

So a few facets to what you mention i think are driven by two historical patterns from Solarwinds as a company. They have had a long history of not coming up with new data, your machines emit this data and SW can display it but we won't recalculate it or come up with new numbers/strings/whatever. Even in cases where there…

I have this query for the Custom Query widget that seems like it should have what you need. https://github.com/Mesverrum/MyPublicWork/blob/master/SWQL/Node%20Downtime%20History

It was only for dvswitches back when I did my vmware certs, and it looks like that's still the case in their docs. techdocs.broadcom.com/.../configure-netflow-settings-with-the-vsphere-web-client.html

I'd lay money down that since this is joining to the events table to see when the node went down the 4 missing UPS have been down so long they don't have an entry on the events table anymore, at least 30 days with default settings. You can probably cover them all by making that inner join to events a left join.

Assuming you use vcenter and have the distributed vswitches available with your license type then your vcenter can generate flow for the traffic coming and going to each VM, regardless of the OS. I've also used the Host Sflow tool in the past to generate flows from windows machines, its a bit annoying because all the…

I dont have a sample query handy, but you should be able to check out the events and filter it to just app down events, and then generate a count of those entries grouped by server name. Probably can do that with the GUI report builder and no SQL/SWQL.

Joining the historical stat tables for CPU, memory, and volumes like that is likely generating a huge amount of rows and I wouldn't be surprised if thats timing out. What tool are you trying to report in, power bi or ssrs or something else? This should be broken apart into separate queries for each historical table to stop…

To force swql to treat a string as an integer you can try totallicenses*1 - utilizedlicenses*1 Might need to experiment with it but that sort of thing tips it if that it's dealing with numbers. 

There are several OOTB reports and widgets showing average availability for nodes, it sounds like you want one of those or at least that would get you close.

To address "has this been done" the answer is yes, it is very common. I've seen this at a couple dozen clients myself. It's pretty trivial, but the biggest obstacle tends to be that the PBI data engineers usually know next to nothing about the kinds of data available in SW so they struggle to figure out the relevant…

what oid are you adding for this license info?

Its almost the first thing at the top of the page when you edit nodes, something about dns or use hostname instead of IP. Youll find it if you check.

MS docs have examples for every built int powershell function https://learn.microsoft.com/en-us/powershell/module/microsoft.powershell.utility/invoke-webrequest?view=powershell-7.5 specifically in a SW context these threads should help SAM Powershell Monitor for cURL check Using SAM to Call a Rest API (Rightfax Faxing job…

Should be similar to select nodeid, caption, ip_address, totallicenses, utilizedlicenses, utilizationpercentag, (totallicenses - utilizedlicenses) as remaininglcenses from ( SELECT n.NodeID, n.Caption, n.IP_Address, MAX(CASE WHEN AssignmentName LIKE '%avAesTsapiLicenseTotal%' THEN CurrentValue END) AS TotalLicenses,…

Since SW is installed primarily in Windows you might opt to use a powershell script monitor with invoke-webrequest. If you must use curl for some reason then you'll need to install it to the server and wrap it in a script to call and parse the results.

There is an option in the edit nodes view to use dns instead of the IP to poll a device. I'd pull up the docs link but apparently the docs page is having trouble tonight

I looked through some of my old scripts and it looks like I had to resort to editing thresholds directly in SQL, so at least the last time i had the need to write this kind of script it could not be done via the API.

I would nest this all into a subquery and then you can do the math operations against the various columns.

Solarwinds gets the information about the speed of an interface from the bandwidth statement, so the best place to update those is to actually set the bandwidth on the devices themselves. In the cases where you need to manually set a bandwidth over ride then this script i wrote a few years ago did the job. Havent tested in…