mrxinu

No worries, now that I know you and he both think the client is a good idea it's now a doubly good idea. =)

Yeah, hmm. You almost need an anti-suppression (do this only if this other condition is also true). Since the alert conditions are evaluated on a per-row basis, the node name can't be X and Y at the same time. What about a negated suppression? Trigger condition: node A is up; suppression condition: node B is down. Then the…

That worked just fine for me.

My pleasure, sir.

Okay, so that's not a formatting error - it's the best way to represent a number that big in that amount of space. Would you prefer a large number of zeros?

What happens when you remove the WHERE clause entirely? Do you get rows?

The ClearAlert verb needs an array of AlertIDs inside the array of alert arguments. Try rewriting the Invoke-SwisVerb line like this: Invoke-SwisVerb $swis Orion.AlertActive ClearAlert @(, [int[]]$alertid)

That certainly looks right to me. Can you find your new volume and pollers in Orion.Volumes and Orion.Pollers? If so, try deleting it and adding it with the web console and see if you get differing values for both.

SolarWinds doesn't publish an ERD (database documentation) because they don't support direct access of the database. So your only alternative is to get into the database and start poking around. I would start with the views because they a) give you immediate access to useful information, and b) offer some insight into…

Here's one I hacked together in vbscript that seems to work well. It takes the computer name and the short name of the process as command line arguments, so in the trigger action it would look like: c:\path\to\script\restart-service.vbs ${Node.IP_Address} ${Component.ServiceName} (I think that's the variable - you'll need…

Absolutely, what @"mesverrum" said. I pushed an example doing exactly what you're looking for to the SolarWinds repository. Until it's merged, you can see it here: https://github.com/solarwinds/OrionSDK/pull/216/commits/ed0a2679fe7aa31280b8c6da1e88d8dc217d37e8

You're welcome. Ah yeah, agreed on the plain-text password - not such a good idea. You could encrypt the password in the custom property and then decrypt it in the vb script possibly. It's not perfect, but at least it would keep the actual password out of a search.

It would be cool, but having been a UI developer in a previous life there's a big difference between kicking out a script that handles a few command line options (recipient, node name, IP address, etc), opening a connection to the XMPP server, and sending the message; and building up a UI to accept all that information,…

The Tomcat components are individually named for what they do so you can just include the component name in the email: -- Steven W. Klassen Programmer Analyst @ Loop1 Systems http://www.loop1systems.com/ http://www.linkedin.com/in/mrxinu

Woohoo! Teamwork. *high-fives aLTeReGo​* -- Steven W. Klassen Programmer Analyst @ Loop1 Systems http://www.loop1systems.com/ http://www.linkedin.com/in/mrxinu

Is "WAN" a default property for Interfaces? I don't see it in the list. No, it's not a default. You would change the radio button to 'build' and then set the type to 'Interface', the name to 'WAN', and the type to 'Yes/No'.

After going to 9.x the system manager was defunct, so if you're on any 9.x release you really ought to be in the web console. The custom properties are still managed (adding & removing them) via the thick client located under SolarWinds Orion > Grouping and Access Control > Custom Property Editor.

How many rows are in that table? Any duplicate NodeIDs? SELECT NodeID, COUNT(1) AS QtyFROM VIM_HostsGROUP BY NodeID

Right on. If you want a hand navigating the GitHub stuff and doing a pull request I'd be happy to hop on a GTM with you and run you through it. Gimme a shout.

Exactly - there's an escalation tab on each action. So if you wanted an email at the outset of the issue, you would create that as usual. The next action (another email?) would have the box checked in the 'Escalation' tab that says "delay execution of this action for X minutes/hours". Same thing with the third action only…

And it appears to happen 7 times when it triggers and 7 when it resets. How bizarre. I'll sleep on it and let you know if I come up with anything in the morning. How many alerts do you have enabled? Is it possible you have more than one definition of the alert and they're overlapping (e.g. maybe you intended to alert on…

My pleasure. I'm stepping away for a couple of hours, but let me chew on this and get back to you when I return.

I haven't used it personally - I talk about the product as part of the training on alerts, but to date I haven't had anyone who wanted it configured while I was on site. Customers have set it up at after I've left and I haven't heard any complaints about it being broken. If I remember what I heard about its configuration,…

Okay, this will work better. We'll use the credentials passed from SAM to the PowerShell script and leverage WMI to get the active drives on the target system. Beyond that, it's the same routine as last time. The screenshot shows that it found two log directories (one on C:\ and one on F:\) with a variable directory just…

Right, gotcha. I'll make the change and post here shortly.

When I have it write the alert message out to a file on my desktop it comes across as "MM/DD/YYYY hh:mm:ss". What's the date format set to on the application server?

Strange. Let me tinker a bit this afternoon to work out the easiest method.

Looks good! Don't forget to mark jorich​'s reply about the java path as the answer so this thread gets converted to complete!

Duh, I totally missed that too. Darn HTML entities.

I would try stripping everything out but the connect, building the $poller hash, and the New-SwisObject call and see how you fare with that. I wonder if it's a scope issue. I don't think I've seen the entire script so I don't know where variables are being created. Also, BRB, need to get 3 hrs of sleep in before my next…