sturdyerde · IT Solutions Team Lead / Microsoft PowerShell MVP · ✭✭✭✭✭

@"gothamyyc" wrote:Thanks for the replies mesverrum, sturdyerde, sreenathmp! I was told that applications and services couldn't be monitored when using WMI, and wondered if that was actually the case. That is incorrect. Applications and services can be monitored with WMI. 

@"MathieuJM" wrote:Do you plan a true RBAC design for the whole products ? it would be nice to have something constant from one SolarWinds product to another one. (admin, operator, read-only, none, ...) For exemple to manage agent we must be administator Plus all of the ones for this question/request!!!!!

Just trying to figure out how/where we can help you with this question.

How did the upgrade end up? Hope it was without incident after this delay!

Ouch, sorry to hear that yours has gone so poorly. Do you suspect .NET 4.8 or something else?

wrote: no issues with the install that required a restart or back out, total install time for my environment was 32 minutes. my production environment consist of 779 Nodes, 7313 Interfaces, 376 Volumes, 8468 Total Elements. single polling engine with NPM\NTA Which version[s] did you upgrade from?

wrote: So does anyone recommend doing the .NET 4.8 install before or let the installer take care of it? I would definitely recommend installing ​and patching​ .NET 4.8 before the upgrade. That will save time during the upgrade and also give you a chance to confirm that your current installation is still healthy after the…

Thanks for the links. Will the contents of these HFs be baked into the latest platform downloads, or will they need to be applied after we upgrade to 19.2?

superfly99 wrote: If you edit an alert, the alert will not re-trigger. If you create (or copy) a rule, then it will trigger on what ever the condition is even if it's the same as an existing rule. In my experience, editing an alert does re-trigger if a new alert condition is met, or if a previous instance of the alert was…

Your "this would become" screenshot of the new alert criteria is missing a condition for "F5 LTM Pool - F5 Status - is equal to - Down". If you fix that, the trigger count should go down, I think.

Yeah, it sounds like you may want to (A) create a separate alert that can filter by client, and (B) create a custom property on nodes to identify the client. The custom property could also be used to route alerts via actions.

I found some stuff, but am going to stop here before I void my and anybody else's warranty. This can just be a feature request unless someone has a supported solution out there!

serena​, tdanner​: is that expected behavior when upgrading to SAM 6.8?

Continuing this quest, I spoke with a developer colleague and he determined that api.opsgenie.com doesn't allow CORS requests (Cross-Origin Resource Sharing). In his words, "Basically, for security reasons the browser prevents javascript code from making requests to a domain that is different than the origin from which it…

Since I'm most familiar with PowerShell, I started by testing with that. The following code successfully pulls the "name" (email address) of the on call person. My next step will still be to figure out how to translate this query into JavaScript. $key = "REDACTED" $header = @{'Authorization' = 'GenieKey ' + $key}…

Thanks. This gives me a little to go on, although I have almost zero experience with scripting API calls. Will have to keep digging and learn as I have time unless someone else has already tried this!

serena​, that is nice to know! Thanks.

christopher.t.jones123​, thanks for including the reference!

We're doing the same with OpsGenie. Our security team is world class and appropriately strict, but they approved the 2-way integration in this case because there are no open ports exposed externally. The integration agent lives on one of our Orion servers and it simply checks a messaging pub/sub service in OpsGenie for new…

I don't see why this wouldn't work. One would just need to create proper JSON structure around the variables and then send it with an HTTP Post action.

wrote: 1. Deploy SAM on the NPM server to use existing NPM database and then install two APEs (which will again connect to same database). 2. You need additional license for APE's (In total you will have your NPM license, SAM license and 2 X Additional poller license). Note: SAM on a seperate orion server with two APEs…

That's odd that they haven't cast any OIDs in the virtual appliances, but then again, I guess the Core and Sentry really are just Linux machines.The odd thing is my previous monitoring platform (EM7) discovered them as Cisco VOIP devices. osborne_graham​, have you created any application monitoring templates for MobileIron…

silverbacksays wrote: Hey hpstech If you're able to use PowerShell to power the query, you can try using Invoke-WebRequest instead. Since you're looking to interrogate a remote system, the method of grabbing the data shouldn't matter, if you're passing a URL to the command. Give it a try and let us know how you get on!…

Thanks for checking, serena​! My Docker engineer here responded that he is using 'root' which should eliminate any questions/problems with permissions.

Reviving this question -- should I infer from your answer that you have added each VIP as a node in NPM and then applied a SAM template to monitor the URL:port on that VIP?

Expanding on what hpstech​ said, it is definitely recommended to monitor Windows computers with WMI instead of SNMP. If you have successfully setup WMI monitoring, you should then disable or remove the SNMP service[s] on your Windows computers.

Hmmm, I wonder what ccousineau​ and serena​ would say to a sort of customer advisory board that could suggest realistic OOB thresholds and also offer input on what metrics to include in OOB templates? (The thought of a voting, forum, or wiki-styled input forum also crossed my mind, but that could get messy and off-topic…

This is a great little tip! Glad I found it. Here's what I did to linkify email addresses in a custom widget. (The ComponentID is irrelevant to any of you...it's simply the specific application monitor component that pulls the email address of whoever is on call for each team here.) SELECT ComponentName AS Team,…

Yeah, it looks like a slip on their usage of element versus component. Thanks for sharing the query!

With my PowerShell monitors, I sometimes have trouble with Orion variables resolving directly in cmdlet parameters. The workaround has been to save the Orion variable to a PowerShell variable at the top of the script, and then reference the PowerShell variable in code. Try using this at the top: $c = Get-Credential…