mesverrum · Observability Architect · ✭✭✭✭✭

I happened to set this up with my client this week and it won't run for them using NPM 12.2, so its possible that this would require 12.3, I'll pin down the version and update the post.

Do you have SWQL studio installed? Try it there and you tend to get much more detailed error messages. Troubleshooting it through the web UI is rarely productive. https://github.com/solarwinds/OrionSDK/releases

I'm running the latest release, can't see anything that jumps out to me as requiring a specific version, which one are you running?

Event history is only kept for 30 days if you have the default retention settings

in the where section add something to the effect of AND n.customproperties.myproperty = 'whatever'

I noticed that there is a mixture of tolocal and utc timestamps being used in that query, you can change the minutes line to use getdate() instead of getutcdate() and it should correct the measurement.

It can be a little tricky since these change depending on the type of object you are alerting on, but try these: Application Name ${N=SwisEntity;M=Application.ApplicationAlert.ApplicationName} Node Status ${N=SwisEntity;M=Application.Node.Status} Application Status, formatted to show as up/down instead of 1/2…

Swql and sql are not interchangeable. Code that works in one will almost guaranteed not work on the other without changes.

Copy pasta error, i just edited that one to fix it.

I use a different, simpler one for just seeing nodes that are currently down SELECT n.caption as Node ,n.detailsurl as [_linkfor_Node] ,'/Orion/images/StatusIcons/Small-' + n.StatusIcon AS [_IconFor_Node] ,n.ip_address as [IP Address] ,n.detailsurl as [_linkfor_IP Address] ,case when n.LASTSYSTEMUPTIMEPOLLUTC is null then…

To validate that a program is doing what it is supposed to you usually need to find a way to test that you can access the app and get some kind of expected response. Some apps are entirely GUI driven which basically leaves you helpless, but anything with any kind of API, CLI, or some other scheme where we could…

Try this select n.customproperties.[yourcustomproperty] , n.caption as [Device] -- shows the current status icon , '/Orion/images/StatusIcons/Small-' + n.StatusIcon AS [_IconFor_Device] -- makes a clickable link to the node details , n.DetailsUrl as [_linkfor_Device] -- shows the timestamp of the down event, if there is no…

To do things like subtotals you would need to union basically two separate versions of the query together, one how I have it, and one with sum of the downtime figures and a bunch of placeholders, since a union requires the same number of columns, but on a summary scenario many of these columns wouldn't be…

NPM automatically picks up all available interfaces when you add a node and then you choose when ones you want to monitor. Run a discovery or hit list resources on the node details of an xisting node and you will see loopbacks wherever they exist. Do you mean switching loops like when STP isn't working correctly?

So this thing is a bit complex so here's how it works. The initial query starts out by checking the events table for event type 1, which is a node down, and aliases this as starttime. Then it uses a subselection that I'm calling endtime, to say "for each starttime you see, check the for the first up event (event type 5)…

The linkfor and iconfor syntax only works with the custom query resource, nothing else will acknowledge them. You can embed the custom query resource in Web based reports, on the screen where it asks if you want to do a table or chart just search for query and it pops up.

Yes possible, it's kind of a pain but here is how I do it with another query I wrote ,CASE when minutediff(o.AlertActive.TriggeredDateTime,GETUTCDATE())>1440 then (tostring(round(minutediff(o.AlertActive.TriggeredDateTime,GETUTCDATE())/1440.0,1)) + ' Days') when minutediff(o.AlertActive.TriggeredDateTime,GETUTCDATE())>60…

Probably best to just change your outlook on that bell icon, it almost never actually displays information about actual system issues, its really just FYI kind of messages.

This query is not meant for the report writer, it doesn't support some of the things I have it doing. Build a report and instead of using the custom table resource look for the custom query resource and paste it into that one instead, then it should work for you.

Times can be a bit annoying depending what exactly you want For something simple like last 30 days you could try adding to the where condition near the end and t2.[down event] > addday(-30,getutcdate()) Getting into previous calendar month is a bit more complex, this is how I'd probably go about that and t2.[down event] >…

In every client environment I've been the contents of that table were wildly inaccurate, so I had to roll my own data.

I'll admit, i love complexity haha, I spent an afternoon with a Linux admin a few weeks ago getting openssh configured on our orion server and then leveraging powershell and openssh that to be able to ssh into their linux machines with a cert instead of user/pass combo and then executing whatever scripts we wanted to cook…

To be accurate you are seeing multiple down events correlating to a single up. There are some glitchy scenarios in Orion where it records downs over and over despite the fact that the node never came up. Haven't come up with a good workaround for that yet.

Those variables will only show you info regarding the current alert. You would probably be better off trying to build a report against the alert histories. It wouldnt go out as the alerts are acknowledged in real time but it can show you everything in one go for the previous day or whatever interval.

In the UNDP tool on the server you can edit the poller and under advanced options change the value type from rate to counter. It will automatically show as the difference between the current and previous polled value https://thwack.solarwinds.com/thread/19684#50453

I'll say that under most use cases solarwinds is pretty overkill for home use, the smallest license size is still several thousand dollars.

Here is a guide to an application that should do what you want How-to configure a free NetFlow forwarder or NetFlow duplicator - Brad Reese

Not as a part the same table as the more granular data, unless you know some SQL wizardry I've never seen before to have a single query generate grouped and ungrouped data inline. For that type of request I tend to make a separate table just above with the aggregated values I want. Not as slick in terms of presentation and…

For an Http/s monitor all testing happens at the Orion server.

That is the expected behavior, Solarwinds does the group matches to the first group you are a member of based on the order listed in the account management page, it does not combine group memberships or anything else like that. Local or individual account permissions also supersede any group memberships so if you have an…