Azure-cli: metrics alert update does not seem to work

[tobiasahlm]

Describe the bug
The update function for metrics (Alerts) does not seem to work.
I have created an alert for an Application Gateway to monitor health. When you try to update the property enabled, it returns a bad response.

To Reproduce
Steps to reproduce the behavior:

az monitor metrics alert update --name testAlert --resource-group testRG --enabled false --debug

msrest.http_logger : Response content:
msrest.http_logger : {"Code":"BadRequest","Message":"targetResourceType property is missing or invalid."}
msrest.exceptions : Operation returned an invalid status code 'BadRequest'

Expected behavior

Expected no error and portal.azure.com would reflect the changes and disable the alert mentioned above.

Environment summary
az--version: azure-cli (2.0.54)
Windows 10 - Powershell

Additional context
I have used the CLI in powershell and https://docs.microsoft.com/en-us/cli/azure/monitor/alert?view=azure-cli-latest#az-monitor-alert-update

It indicates how I used it is supposed to be correct.

[tjprescott]

@tobiasahlm could you please provide the commands you used to create the application gateway and the alert itself?

[tobiasahlm]

@tobiasahlm could you please provide the commands you used to create the application gateway and the alert itself?

@tjprescott The application gateway and the alert was created through the Azure portal. I will try to create a scenario with CLI commands and let you know the outcome.

[tobiasahlm]

I have investigated it further and have the following findings:
1) It does not matter if the application gateway is created through the Azure portal or CLI.
2) I can enable and disable alerts if I created it with the following CLI command:
az monitor metrics alert create --name TempTest --resource-group AppGWTest --condition "avg UnHealthyHostCount > 1" --description "Unhealthy!" --scopes {scope} --action {actionGroup}

However if I create the identical alert on portal.azure.com, I am not able to enable/disable the alert. Then I get the invalid property.

I ran az monitor metrics alert list --resource-group appGWTest to get the output of the two alerts. They look similar and the only difference I can see that is added from portal.azure.com is: "webhookProperties": null

`

This one works (I can enable / disable through the CLI)

{
"actions": [
{
"actionGroupId": "ZZZ",
"webhookProperties": null
}
],
"autoMitigate": null,
"criteria": {
"additionalProperties": null,
"allOf": [
{
"dimensions": [],
"metricName": "UnHealthyHostCount",
"metricNamespace": "microsoft.network/applicationgateways",
"name": "cond0",
"operator": "GreaterThan",
"threshold": 1.0,
"timeAggregation": "Average"
}
],
"odatatype": "Microsoft.Azure.Monitor.SingleResourceMultipleMetricCriteria"
},
"description": "Unhealthy",
"enabled": true,
"evaluationFrequency": "0:01:00",
"id": "ZZZ",
"lastUpdatedTime": null,
"location": "global",
"name": "ToahTemp4",
"resourceGroup": "AppGWTest",
"scopes": [
"ZZZ"
],
"severity": 2,
"tags": null,
"type": "Microsoft.Insights/metricAlerts",
"windowSize": "0:05:00"
}

---

## This one does not work (I can't enable / disable through the CLI)

{
"actions": [
{
"actionGroupId": "XXX",
"webHookProperties": {},
"webhookProperties": null
}
],
"autoMitigate": true,
"criteria": {
"additionalProperties": null,
"allOf": [
{
"dimensions": [],
"metricName": "UnhealthyHostCount",
"metricNamespace": "Microsoft.Network/applicationGateways",
"monitorTemplateType": 8,
"name": "Metric1",
"operator": "GreaterThan",
"threshold": 1.0,
"timeAggregation": "Average"
}
],
"odatatype": "Microsoft.Azure.Monitor.SingleResourceMultipleMetricCriteria"
},
"description": "Test",
"enabled": true,
"evaluationFrequency": "0:01:00",
"id": "XXX",
"lastUpdatedTime": null,
"location": "global",
"name": "TempTest",
"resourceGroup": "AppGWTest",
"scopes": [
"XXX"
],
"severity": 3,
"tags": {},
"type": "Microsoft.Insights/metricAlerts",
"windowSize": "0:05:00"
},
`

[tjprescott]

@tobiasahlm thank you for the detailed info. Just to confirm, the second alert is the one you originally created in the portal and are trying to disable via CLI?

[tobiasahlm]

@tjprescott Exactly. The second alert is created through the Azure portal and I am not able to disable it through the CLI. I suppose you will encounter the same issue, if you create an alert on the application gateway through the azure portal.

[asos-wesleykey]

Seeing the same behaviour for alerts that have been created via the portal

[tjprescott]

I can reproduce this behavior. I will note that I can correctly create an alert with webhook properties in the CLI, but if I then update that object, the webhook properties are lost. I believe this is an issue with the REST API specs definition that the CLI depends on.

I have created an issue there.
https://github.com/Azure/azure-rest-api-specs/issues/5065

[tjprescott]

At the same time, I would suggest opening a support ticket in the Azure portal as that will more quickly draw the attention of the service team to this issue.

[osariel]

Hiya, I have just been battling the same error. Any updates or work around?

[tjprescott]

No updates. I would recommend opening a support ticket in the portal.

[tobiasahlm]

I already did that and the customer service told me they would write when they have an update, but it is currently not on their board to fix it. So I hope more will report this, so it become a priority.

[tjprescott]

@tobiasahlm I agree.

[abkasera]

I have tested with the metric alert on the Appilcation Gateway and updated it from the cli. It is working fine now. I'm using cli version(2.0.69)
CLI Command: az monitor metrics alert update --name testgateway --resource-group ABKASERA-RG --description abcxyz --debug

Response payload:
{
"actions": [
{
"actionGroupId": "/subscriptions/a2d51762-a47c-4358-9048-e04041599027/resourcegroups/abkasera-rg/providers/microsoft.insights/actiongroups/testactiongroup",
"webhookProperties": null
}
],
"autoMitigate": true,
"criteria": {
"additionalProperties": null,
"allOf": [
{
"dimensions": [],
"metricName": "Throughput",
"metricNamespace": "Microsoft.Network/applicationGateways",
"name": "Metric1",
"operator": "GreaterThan",
"threshold": 100.0,
"timeAggregation": "Total"
}
],
"odatatype": "Microsoft.Azure.Monitor.SingleResourceMultipleMetricCriteria"
},
"description": "abcxyz",
"enabled": false,
"evaluationFrequency": "0:01:00",
"id": "/subscriptions/a2d51762-a47c-4358-9048-e04041599027/resourceGroups/ABKASERA-RG/providers/Microsoft.Insights/metricAlerts/testgateway",
"lastUpdatedTime": null,
"location": "global",
"name": "testgateway",
"resourceGroup": "ABKASERA-RG",
"scopes": [
"/subscriptions/a2d51762-a47c-4358-9048-e04041599027/resourceGroups/ABKASERA-RG/providers/Microsoft.Network/applicationGateways/testappgateway"
],
"severity": 3,
"tags": {},
"type": "Microsoft.Insights/metricAlerts",
"windowSize": "0:05:00"
}

[snehithm]

@tobiasahlm can you let us know if you are still seeing this issue? We have not been able to repro this with the latest cli version(2.0.69)

[tobiasahlm]

@snehitm - It seems to work now, thanks for the update.

[snehithm]

Thanks, @tobiasahlm. Good to know the issue is resolved and sorry for the inconvenience in the past.