I have following setup with service fabric cluster:
- 5 nodes(Durability tier: Bronze and Reliability tier: Silver)
- Current average CPU utilization % on the related VM Scale Set (VMSS) averages around 14%.
- Services are deployed properly each with instance count of -1.
- Our application that uses the above services runs fine.
Using the azure portal, I attempted to setup auto scaling option for the above VMSS (in Scaling option):
- Used the scale out with a threshold of CPU utilization % as 10 (From above, the current value is 14).
- Configured it to wait for 5 minutes before scaling up.
- Rest of the details are shown in below json.
{
"location": "eastus",
"tags": {
"$type": "Microsoft.WindowsAzure.Management.Common.Storage.CasePreservedDictionary, Microsoft.WindowsAzure.Management.Common.Storage"
},
"properties": {
"name": "My Testing",
"enabled": true,
"targetResourceUri": "/subscriptions/MySubScriptionID/resourceGroups/My-RG-Name/providers/Microsoft.Compute/virtualMachineScaleSets/nt1vm",
"profiles": [
{
"name": "Auto created scale condition 1",
"capacity": {
"minimum": "5",
"maximum": "10",
"default": "5"
},
"rules": [
{
"scaleAction": {
"direction": "Decrease",
"type": "ChangeCount",
"value": "1",
"cooldown": "PT5M"
},
"metricTrigger": {
"metricName": "Percentage CPU",
"metricNamespace": "",
"metricResourceUri": "/subscriptions//MySubScriptionID//resourceGroups/My-RG-Name/providers/Microsoft.Compute/virtualMachineScaleSets/nt1vm",
"operator": "LessThan",
"statistic": "Average",
"threshold": 5,
"timeAggregation": "Average",
"timeGrain": "PT1M",
"timeWindow": "PT5M"
}
}
],
"fixedDate": {
"timeZone": "Mountain Standard Time",
"start": "2018-09-18T00:00:00.000Z",
"end": "2018-09-21T23:59:00.000Z"
}
},
{
"name": "Auto created scale condition",
"capacity": {
"minimum": "5",
"maximum": "10",
"default": "5"
},
"rules": [
{
"scaleAction": {
"direction": "Increase",
"type": "ChangeCount",
"value": "1",
"cooldown": "PT15M"
},
"metricTrigger": {
"metricName": "Percentage CPU",
"metricNamespace": "",
"metricResourceUri": "/subscriptions/MySubScriptionID/resourceGroups/My-RG-Name/providers/Microsoft.Compute/virtualMachineScaleSets/nt1vm",
"operator": "GreaterThanOrEqual",
"statistic": "Average",
"threshold": 10,
"timeAggregation": "Average",
"timeGrain": "PT1M",
"timeWindow": "PT5M"
}
}
]
}
],
"notifications": [
{
"operation": "Scale",
"email": {
"sendToSubscriptionAdministrator": false,
"sendToSubscriptionCoAdministrators": false,
"customEmails": [
""
]
},
"webhooks": []
}
],
"targetResourceLocation": "eastus"
},
"id": "/subscriptions/MySubScriptionID/resourceGroups/My-RG-Name/providers/microsoft.insights/autoscalesettings/My Testing",
"name": "My Testing",
"type": "Microsoft.Insights/autoscaleSettings"
}
Observation:
- I waited more than 5 minutes (which is timeWindow value for scale out option). No new nodes are created.
- No new nodes are created after waiting for 10 minutes, 20 minutes and 30 minutes.
- When I configure the number of nodes to 6 (from 5) in the azure portal in the above VMSS (in Scaling option) manually, the node is created and everything work as expected.
Question:
- What am I doing wrong in the above configuration for auto scaling option?
