mesverrum · Observability Architect · ✭✭✭✭✭

Ah yeah, now that I'm at my desk at work I can see, you are averaging the historical data for the entire history of every interface every time you pull a sample. That's not really productive for a chart like you are trying to build. You should just be pulling the point in time values for inbps and outbps from…

Its just adding it up on the fly.

Depending how you quantify failure. For example I don't generally bother to do HA on APE's because I wrote a set of scripts I like with logic that will move all nodes off a poller if it stops working and I just shuffle them to a different polling engine that's still operational. As long as you have n+1 spare capacity then…

You just need to join all those separate select statements on rowid's. Here is an example of a SWQL table I built a while ago doing something similar for a stack of custom F5 UNDP pollers before solarwinds did the native integration. Sorry it is such a complex example but hopefully you can see how to apply the logic to…

Your original post said just 2019.1, haven't run into the problem you describe.

No.

On that note, vote it up https://thwack.solarwinds.com/t5/NPM-Feature-Requests/Change-Perfstack-chart-behavior-for-missing-data/idi-p/557248

This series of posts is the best jumping in point for SW scripts https://thwack.solarwinds.com/t5/Geek-Speak-Blogs/The-Basics-of-PowerShell-part-3/ba-p/434046

You can't remove a required property. If you made the property not required then you can just blank it out.

Depending how old your orion db is it would either be netperfmon or solarwindsorion, since you see netperfmon then that is likely the one you need. I suspect you are right about it being a permissions issue, you'll just want to have your dba assign the necessary permissions to the account the splunk tool is using to pull…

That's a legit point, the undp gets you away from the dependency on monitoring the interface. On your switches are you able to get the local port Id column then you could join across the two tables with those.

What is your interface statistic polling interval and what is the time range set on the chart? SNMP bps calculations are actually an average, they poll the counters for total bits sent/received and then divide that by the amount of time between the current and previous polling interval, so it never really accurately…

looks like you set up a process monitor, the article is telling you to set up a perfmon monitor looking at the perfmon tracking the number of processes that are running. Go back to the wizard and set up a perfmon counter as it explains and you should be better.

My mistake, was going from memory they are actually in containermemberdefinition This SWQL query gets you the DQ's themselves SELECT TOP 1000 DefinitionID, ContainerID, Name, Entity, FromClause, Expression, Definition FROM Orion.ContainerMemberDefinition where definition like 'filter%'

Agent uses less resources on the network, and in theory less resources on the host, but there have been bugs here and there where the agents go haywire and use up all the available resources on the host

At a glance it looks like you might need to specify a rowid. Not in front of my computer to confirm right now though.

There is an out of the box alert called "Restart a Service" that you can copy as your example. On the trigger definition you set up your rules, and the main change is on the actions to this: APM\APMServiceControl.exe ${N=SwisEntity;M=ComponentAlert.ComponentID} and change it to (assuming that 1234 is the componentid of…

This thread goes through some examples in pretty significant detail https://thwack.solarwinds.com/t5/NPM-Discussions/Alert-Event-from-Syslog-or-Trap/m-p/280531

This is the master list of MS IP address, just create an NTA IP Address group with all of these IP's in it…

On step 13 of that KB they list out the oids they use to find the values in the default poller, the only tricky one is the Vendor because they normally do that based on a list they have mapping sysobjectid to vendors.

I'd doubt anyone has a design template to offer you, you'll have to do that leg work yourself.

on your orion web console customize a view and add a resource called the "Custom Query" I typically build out an admin page full of queries like this so i can keep track of who is making what changes in my environment and other issues I need to stay on top of.

This was pretty much the main topic of my SWUG presentation in NYC a few months ago, so I uploaded some sample code to github for that The first script takes a set of node and SAM application custom properties called applications and applicationsrole and builds a tree of groups for every application that currently exists…

It's been a while since I did something like that but I did install a module only on the main poller before. If you try to use a feature from that module on a node that didn't have the module installed it caused errors. If you were strategic you could probably work around it but I think it could get a bit messy. I…

Google cloud platform

I imagine you can't downsample it because it's not a time series, its just a single point in time. Like i said there is no table that actually records that value historically, they just take the count of the current number of sessions and update the current stat each polling interval. In order to calculate what the…

Trial and error? I never have seen a convenient reverse engineering method

This may work for you SELECT [data].[OrionNode].[DisplayName] AS [DisplayName],[data].[RemoteAccessSessions].[UserName] AS [UserName],toLocal([data].[RemoteAccessSessions].[ConnectedTime]) AS [ConnectedTime],toLocal([data].[RemoteAccessSessions].[DisconnectedTime]) AS…

Try Orion.APM.WindowsEvent

You could possibly build a sam script that calls your other existing script and then searches it's response and returns a formatted error count and messsage that sam would understand. Basically just a wrapper around the existing script.