Disclaimer: This is for folks who are running services on Windows machines and does not have more than one device. I am neither an expert at self hosting nor PowerShell. I curated most of this code by doing a lot of "Google-ing" and testing over the years. Feel free to correct any mistakes I have in the code.
Background
TLDR: Windows user needs an uptime monitoring solution
Whenever I searched for uptime monitoring apps, most of the ones that showed up were either hosted on Linux or containers and all I wanted was a a simple exe installation file for some app that will send me alerts when a service or the computer was down. Unfortunately, I couldn't find anything. If you know one, feel free to recommend them.
To get uptime monitoring on Windows, I had to turn to scripting along with a hosted solution (because you shouldn't host the monitoring service on the same device as where your apps are running in case the machine goes down). I searched and tested a lot of code to finally end up with the following.
Now, I have services running on both Windows and Linux and I use Uptime Kuma and the following code for monitoring. But, for people who are still on Windows and haven't made the jump to Linux/containers, you could use these scripts to monitor your services with the same device.
Solution
TLDR: A PowerShell script would check the services/processes/URLs/ports and ping the hosted solution to send out notification.
What I came up with is a PowerShell script that would run every 5 minutes (your preference) using Windows Task Scheduler to check if a Service/Process/URL/Port is up or down and send a ping to Healthchecks.io accordingly.
Import-CSV C:\Services.csv | foreach{
Write-Output ""
Write-Output $($_.Service)
Write-Output "------------------------"
$RESPONSE = Test-Connection localhost -TcpPort $($_.Port)
if ($RESPONSE -eq "True") {
Write-Host "$($_.Service) is running"
curl $HC$($_.Service)
} else {
Write-Host "$($_.Service) is not running"
curl $HC$($_.Service)/fail
}
}
The script looks through the Services.csv file (Line 1) and check if each of those ports are listening ($($_.Port) on Line 5) and pings Healthchecks.io (Line 8 or 11) based on their status with their appropriate name ($($_.Service)). If the port is not listening, it will ping the URL with a trailing /fail (Line 11) to indicate it is down.
Service
The following code is to check if a service is running.
You can add more services on line 1 in comma separated values. Ex: @("bazarr","flaresolverr")
This also needs to match the Slug.
$SERVICES = @("bazarr")
foreach($SERVICE in $SERVICES) {
Write-Output ""
Write-Output $SERVICE
Write-Output "------------------------"
$RESPONSE = Get-Service $SERVICE | Select-Object Status
if ($RESPONSE.Status -eq "Running") {
Write-Host "$SERVICE is running"
curl $HC$SERVICE
} else {
Write-Host "$SERVICE is not running"
curl $HC$SERVICE/fail
}
}
The script looks through the list of services (Line 1) and check if each of those are running (Line 6) and pings Healthchecks.io based on their status.
Process
The following code is to check if a process is running.
Line 1 needs to match their Slug
$PROCESSES = @("tautulli","jellyfin")
foreach($PROCESS in $PROCESSES) {
Write-Output ""
Write-Output $PROCESS
Write-Output "------------------------"
$RESPONSE = Get-Process -Name $PROCESS -ErrorAction SilentlyContinue
if ($RESPONSE -eq $null) {
# Write-Host "$PROCESS is not running"
curl $HC$PROCESS/fail
} else {
# Write-Host "$PROCESS is running"
curl $HC$PROCESS
}
}
Import-CSV C:\Servers.csv | foreach{
Write-Output ""
Write-Output $($_.Server)
Write-Output "------------------------"
$RESPONSE = Test-Connection $($_.IP) -Count 1 | Select-Object Status
if ($RESPONSE.Status -eq "Success") {
# Write-Host "$($_.Server) is running"
curl $HC$($_.Server)
} else {
# Write-Host "$($_.Server) is not running"
curl $HC$($_.Server)/fail
}
}
Task Scheduler
For the script to execute in intervals, you need to create a scheduled task.
Open Task Scheduler, navigate to the Library, and click on Create Task on the right
Give it a name. Ex: Healthcheck
Choose Run whether user is logged on or not
Choose Hidden if needed
On Triggers tab, click on New
Choose On a schedule
Choose One time and select an older date than your current date
Select Repeat task every and choose the desired time and duration. Ex: 5 minutes indefinitely
Select Enabled
On Actions tab, click on New
Choose Start a program
Add the path to PowerShell 7 in Program: "C:\Program Files\PowerShell\7\pwsh.exe"
Point to the script in arguments: -windowstyle hidden -NoProfile -NoLogo -NonInteractive -ExecutionPolicy Bypass -File C:\healthcheck.ps1
Rest of the tabs, you can choose whatever is appropriate for you.
Hit Ok/Apply and exit
Notification Method
Depending on the integration you chose, set it up using the Healthchecks docs.
I am using Telegram with the following configuration:
Name: Telegram
Execute on "down" events: POST https://api.telegram.org/bot<ID>/sendMessage
Request Body:
```
{
"chat_id": "<CHAT ID>",
"text": "đź”´ $NAME is DOWN",
"parse_mode": "HTML",
"no_webpage": true
}
```
Request Headers: Content-Type: application/json
Execute on "up" events: POST https://api.telegram.org/bot<ID>/sendMessage
Request Body:
```
{
"chat_id": "<CHAT ID>",
"text": "🟢 $NAME is UP",
"parse_mode": "HTML",
"no_webpage": true
}
```
Request Headers: Content-Type: application/json
Closing
You can monitor up to 20 services for free. You can also selfhost Healthchecks instance (wouldn't recommend if you only have one machine).
I've been wanting to give something back to the community for a while. I hope this is useful to some of you. Please let me know if you have any questions or suggestions. Thank you for reading!
Quite a lot of work went into your program, I bet you learned some good code while writing it. I've used a simple Windows program called Easy Net Monitor. You can specify IP address or hostname connectivity tracking with email alerts.
This app would be useful if you have more than one device. If you run the monitoring application on the same device as the services you are monitoring and if the device goes down, you wouldn't get a notification, right?
Correct, if that machine goes down you won't have any notification. I was thinking along the lines of you are hosting services on one machine while the computer itself is acting as the host or hypervisor.