Azure Alerts are not triggered with scheduledTransferPeriod 2M for Diagnostics - Portal (RM)

Hello,

In order to put alerts for a VM we should enable diagnostics (metrics) for it  and then create the desired alerts.

Now, I have some problems with triggering an alert for CPU guest usage when I have enabled this type of diagnostics with this json code : 

{"name": "agent","properties": {"name": "agent","publicConfiguration": {"odata.type": "Microsoft.Azure.Management.Insights.Models.PublicMonitoringConfiguration","diagnosticMonitorConfiguration": {"overallQuotaInMB": 4096,"metrics": {"resourceId": "/subscriptions/xxxxxxxxxx/resourceGroups/xxxxx/providers/Microsoft.Compute/virtualMachines/xxx","aggregations": [
            {"scheduledTransferPeriod": "PT1H"
            },
            {"scheduledTransferPeriod": "PT5M"
            }
          ]
        },"diagnosticInfrastructureLogs": {"scheduledTransferLogLevelFilter": "Warning","scheduledTransferPeriod": "PT5M"
        },"directories": {"iisLogs": "vhds","scheduledTransferPeriod": "PT5M"
        },"performanceCounters": {"counters": [
            {"annotations": [
                {"value": "CPU percentage guest OS","locale": "en-us"
                }
              ],"counterSpecifier": "\\Processor\\PercentProcessorTime","sampleRate": "PT15S","unit": "Percent"
            }
          ],"scheduledTransferPeriod": "PT0S"
        },
      },"storageAccount": "xxxx"
    }
  }
}

With an alert created like this : 

{"location": "xxxx","tags": { },"properties": {"name": "CPU Idle","description": "CPU Idle Time","isEnabled": true,"condition": {"odata.type": "Microsoft.Azure.Management.Insights.Models.ThresholdRuleCondition","dataSource": {"odata.type": "Microsoft.Azure.Management.Insights.Models.RuleMetricDataSource","resourceUri": "/subscriptions/xxxxxxxxx/resourceGroups/xxx/providers/Microsoft.Compute/virtualMachines/xxx","metricName": "\\Processor\\PercentProcessorTime"
      },"threshold": 10,"windowSize": "PT8H","operator": "LessThan","timeAggregation": "Maximum"
    },"action": {"$type": "Microsoft.WindowsAzure.Management.Monitoring.Alerts.Models.RuleEmailAction, Microsoft.WindowsAzure.Management.Mon.Client","odata.type": "Microsoft.Azure.Management.Insights.Models.RuleEmailAction","sendToServiceOwners": false
    },"actions": [
      {"$type": "Microsoft.WindowsAzure.Management.Monitoring.Alerts.Models.RuleWebhookAction, Microsoft.WindowsAzure.Management.Mon.Client","odata.type": "Microsoft.Azure.Management.Insights.Models.RuleWebhookAction","serviceUri": "xxxxxxxxx","properties": {}
      }
    ]
  }
}

The problem comes from bold text when I enable the diagnostics because it's telling me in portal "No available data" and nothing happens with my alert.  The alert is not triggered also for other scheduledTranferPerioad values, starting from 2 minutes and  > .

"scheduledTransferPeriod": "PT5M"

If I use 1 minute sample rate, everything goes  ALL RIGHT and it's working! 

"scheduledTransferPeriod": "PT1M"

Why I need a lower sample rate? Because I want to increase my  alert windowsize  *24 hours

I need to use a slow sampling rate because I want to put my alert to check it for  the last 24 hours and IF I am using 1 minute sample rate for metrics I am over the maximum data points (1000).

So, using 1 minute sample rate for metrics I can use my alert for maxim 16hours window [ 60 (minutes) / 1 (minute sample rate)  * 16 = 960 < 1000 which is OK ]

For 24hours  alert window size and 1 minute sample rate  I will have 60/1 * 24 = 1440 > 1000  (It triggers me an error) 

So I should have at least 2 minute sample rate ! For this sample rate I will have 60/2 * 24 = 720 < 1000 (GOOD) 

please advise.

Thanks!