HA aware scheduled unmanaging/muting of devices, via the Orion interface.

Currently, when i schedule a server to be unmanaged, i need to create the task, then schedule it on my primary server, then copy the task over and reschedule it on my secondary, in order for it to work when I fail over my primary poller.

I would be great if the unmanageing and now the muting of alerts could be done via the Orion console and not using local applications and multiple tasks schedulers.

We can schedule a single downtime from Orion, any chance this could be extended to allow repeat schedules, for un-managing and muting alerts please?

Find more posts tagged with

orion ha

unmanage node

scheduling utility

Status: None

Comments

ryan.davis26

Totally bumping this one!!

stuartd

We've just come across this issue as well so bumped.

I have to say, I am NOT impressed by a lo ot Solarwinds responses which say - create a request and we;ll implement it. So far, two key things we need are 6yrs and 16mths in the waiting with no answer from SW as to a likely timescale for when they will.

yaquaholiccan I ask how you managed to get the schedule copied over as when we try we get an error on the HA box saying something like - "Unable to connect to Orion server. There was no endpoint listening at ..." ??

dsimpkins

I feel your pain on this one but until SW do something about it i had to think out of the box.

Here is an approach that i took and can be run from anywhere you like....

Use PowerShell to talk to Orion and do the unmanaging instead..

Unmanaging Entities · solarwinds/OrionSDK Wiki · GitHub

PowerOrion A Module for PowerShell · solarwinds/OrionSDK Wiki · GitHub

Get yourself a VM outside of your Orion installation and install the PowerShell module, create a .ps1 file with the example and then have a scheduled task fire it when you want it. here comes the problem of which server is active...

I have this PowerShell function, you give it both servers $OrionServer and $SecondaryOrionServer . It will connect to the first and try and make a query, if it fails it tries the second and then hopefully connects with a nice message.

Function GTS-Orion-ConnectToSwis

{

Param(

[Parameter(Mandatory=$true)]

[Validatenotnullorempty()]

$OrionServer,

[Parameter(Mandatory=$false)]

[Validatenotnullorempty()]

$SecondaryOrionServer,

[Parameter(Mandatory=$true)]

[Validatenotnullorempty()]

$Credentials)

# Make connection to SWIS and test with a simple query

$SWISConnectionTest = "

SELECT Name, InstanceType, InstanceSiteID

FROM Orion.SwisFeature

$Swis = Connect-Swis -Hostname $OrionServer -Credential $Credentials

Try

{

$SWISConnectionTestPrimary = Get-SwisData -SwisConnection $Swis -Query $SWISConnectionTest

}

Catch {}

If (!($SWISConnectionTestPrimary))

{

If (!($SecondaryOrionServer))

{

Write-Host 'Unable to Connect to Orion Services' -ForegroundColor Red -BackgroundColor Black

}

Else

{

$Swis = Connect-Swis -Hostname $SecondaryOrionServer -Credential $Credentials

Try

{

$SWISConnectionTestSecondary = Get-SwisData -SwisConnection $Swis -Query $SWISConnectionTest

}

Catch {}

If (!($SWISConnectionTestSecondary))

{

Write-Host 'Unable to Connect to Orion Services' -ForegroundColor Red -BackgroundColor Black

}

Else

{

Write-Host 'Connected to Secondary Server' -ForegroundColor Green

}

Else

{

Write-Host 'Connected to Primary Server' -ForegroundColor Green

}

Return $Swis

}

Want to make your VM running the jobs more resillient? create a second and install the failover clustering on both and create an empty role, use PowerShell and the Get-ClusterGroup cmdlet to return the owner node. run the same scheduled task on both VM's but have it use the cmdlet to find if the local server is the owner, If it's the owner it runs otherwise it quits until the next time.

Yes yes yes SW should make this easier but until then i have something i control end to end and have made much more powerful than the unmanage utility will ever be for me.

yaquaholic

Manually....

I created the schedule on one of the Orion servers, using Windows Scheduled Tasks, then exported it to a file, manually copying it over to each server, with the relevant unmange script.

Repeating the process for each scheduled task.