#Requires -Version 5.1
<#
    JbTecWiz Support Centre -- generated fix script

    Fault : A clustered role will not come online or fails over repeatedly
    Fix   : Find the resource at the bottom of the dependency chain
    Source: https://jbtecwiz.com/support/srv-clu-role-failed

    Run as     : Elevated PowerShell
    Expect     : 50 minutes
    Risk       : medium
    Reversible : yes

    WHEN THIS IS THE RIGHT FIX
    A role fails to start or keeps failing over.

    HOW TO UNDO IT
    Restore the RestartAction and preferred owners once the cause is fixed,
    or the role will not fail over when it genuinely should.

    This script walks the fix one step at a time and asks before each
    one. Steps with no command are things you do yourself -- it prints
    those and waits. Run with -DryRun to print without executing.

    --------------------------------------------------------------------
    NO WARRANTY - USE AT YOUR OWN RISK

    This script is provided by JbTecWiz as-is and with no warranty of any
    kind, express or implied. You run it entirely at your own risk.

    JbTecWiz accepts no liability for any loss or damage arising from its
    use, including but not limited to data loss, downtime, or configuration
    changes that turn out to be wrong for your system.

    You are responsible for reading this script before running it, for
    satisfying yourself that it suits the machine in front of you, and for
    having a working backup first. Some steps cannot be undone.
    --------------------------------------------------------------------
#>

[CmdletBinding()]
param(
    # Print every step and command without running anything.
    [switch]$DryRun,
    # Do not ask before each step. Read the script first if you use this.
    [switch]$Unattended
)

$ErrorActionPreference = 'Stop'
$script:Failed = 0

function Write-Rule { param([string]$Text)
    Write-Host ''
    Write-Host ('-' * 70) -ForegroundColor DarkGray
    if ($Text) { Write-Host $Text -ForegroundColor Cyan }
}

function Show-Prose { param([string]$Text, [string]$Colour = "Gray")
    if (-not $Text) { return }
    $words = $Text -split "\s+"; $line = "  "
    foreach ($w in $words) {
        if (($line.Length + $w.Length) -gt 74) { Write-Host $line -ForegroundColor $Colour; $line = "  " }
        $line += "$w "
    }
    if ($line.Trim()) { Write-Host $line -ForegroundColor $Colour }
}

function Invoke-Step {
    param(
        [int]$Number, [string]$Do, [string]$Why,
        [scriptblock]$Command, [switch]$Manual, [string]$Shell = "powershell"
    )
    Write-Rule "  Step $Number of 6"
    Show-Prose $Do "White"
    if ($Why) { Write-Host ""; Show-Prose $Why "DarkGray" }

    if ($Manual) {
        Write-Host ''
        Write-Host '  -> Do this yourself, then press Enter to carry on.' -ForegroundColor Yellow
        if (-not $Unattended -and -not $DryRun) { [void](Read-Host) }
        return
    }

    Write-Host ''
    foreach ($l in ($Command.ToString().Trim() -split "`n")) {
        Write-Host ("    " + $l.Trim()) -ForegroundColor Green
    }
    Write-Host ''

    if ($DryRun) { Write-Host "  (dry run -- not executed)" -ForegroundColor DarkGray; return }
    if (-not $Unattended) {
        $a = Read-Host "  Run this step? [Y]es / [S]kip / [Q]uit"
        if ($a -match "^[Qq]") { Write-Host "  Stopped at your request."; exit 0 }
        if ($a -match "^[Ss]") { Write-Host "  Skipped." -ForegroundColor DarkGray; return }
    }
    try { & $Command }
    catch {
        $script:Failed++
        Write-Host ("  Step $Number failed: " + $_.Exception.Message) -ForegroundColor Red
        Show-Prose "The rest of the fix may depend on this. Read the write-up before carrying on." "Red"
        if (-not $Unattended) {
            $c = Read-Host "  Carry on anyway? [y/N]"
            if ($c -notmatch "^[Yy]") { exit 1 }
        }
    }
}

Write-Rule
Write-Host '  A clustered role will not come online or fails over repeatedly' -ForegroundColor White
Write-Host '  Find the resource at the bottom of the dependency chain' -ForegroundColor Cyan
Write-Host ''
Write-Host '  Risk: medium   Reversible   50 minutes'
Write-Host ''
Show-Prose 'No warranty. Use at your own risk - JbTecWiz accepts no liability. Read it before you run it, and have a backup.' 'DarkYellow'
Write-Rule

if (-not $Unattended -and -not $DryRun) {
    $go = Read-Host '  Ready? [y/N]'
    if ($go -notmatch "^[Yy]") { Write-Host "  Nothing was changed."; exit 0 }
}

Invoke-Step -Number 1 -Do 'Find the resource that is actually failing, not the role that reports it.' -Command {
    Get-ClusterResource | Where-Object State -ne 'Online' | Format-Table Name,State,OwnerGroup,ResourceType
    Get-ClusterGroup | Format-Table Name,State,OwnerNode
}

Invoke-Step -Number 2 -Do 'Read the dependency chain -- a resource cannot come online until everything it depends on is up.' -Why 'The role reports the topmost resource as failed, but the cause is usually the bottom of the chain -- a disk that did not attach or an IP address that could not be registered. The report makes the order explicit.' -Command {
    Get-ClusterResource 'SQL Server' | Get-ClusterResourceDependency
    Get-ClusterResourceDependencyReport -Group 'SQL Group' -Path C:\
}

Invoke-Step -Number 3 -Do 'Read event 1069, which names the resource and the reason.' -Command {
    Get-WinEvent -LogName System -MaxEvents 100 | Where-Object Id -in 1069,1205,1254 | Format-List TimeCreated,Id,Message
}

Invoke-Step -Number 4 -Do 'Stop the failover loop while you work, so the cluster stops moving it.' -Command {
    Get-ClusterGroup 'SQL Group' | Set-ClusterOwnerNode -Owners node01
    (Get-ClusterResource 'SQL Server').RestartAction = 0
}

Invoke-Step -Number 5 -Do 'Bring the resources up one at a time from the bottom of the chain, so the first failure is obvious.' -Command {
    Start-ClusterResource -Name 'Cluster Disk 2'
    Start-ClusterResource -Name 'SQL IP Address'
    Start-ClusterResource -Name 'SQL Network Name'
}

Invoke-Step -Number 6 -Do 'Get the detailed cluster log for the failure window.' -Command {
    Get-ClusterLog -TimeSpan 15 -Destination C:\clusterlogs -UseLocalTime
}

Write-Rule "  Confirm it worked"
Show-Prose 'The role comes online and stays on one node under normal conditions.' 'White'
Write-Host ''
if (-not $DryRun) {
    Get-ClusterGroup | Format-Table Name,State,OwnerNode
    Get-ClusterResource | Format-Table Name,State,OwnerGroup
}

Write-Rule
if ($script:Failed -gt 0) {
    Write-Host ("  Finished with " + $script:Failed + " failed step(s).") -ForegroundColor Yellow
    Show-Prose 'Read the full write-up at https://jbtecwiz.com/support/srv-clu-role-failed' 'Yellow'
} else {
    Write-Host '  Finished.' -ForegroundColor Green
}
Write-Host ''
Show-Prose 'To undo: Restore the RestartAction and preferred owners once the cause is fixed, or the role will not fail over when it genuinely should.' 'DarkGray'
Write-Rule