Metrics query working on SDK 5.01 now fails wit 6.0

- **Package Name**: python-azure-mgmt-monitor
- **Package Version**: 6.0
- **Operating System**:
- **Python Version**: 3.8.10

**Describe the bug**
python queries worked as expected - Using notebook with Databricks, azsdk-python-azure-mgmt-monitor/5.0.1 Python/3.8.10 (Linux-5.4.0-1100-azure-x86_64-with-glibc2.29)

query fails with http 400 bad request error - Using notebook with Databricks, "userAgent": azsdk-python-azure-mgmt-monitor/6.0.0 Python/3.8.10 (Linux-5.4.0-1100-azure-x86_64-with-glibc2.29)

**To Reproduce**
see URLS below

**Expected behavior**
metric data returned.

**Additional context**
Bad one:
"PreciseTimeStamp": 2023-03-21T04:00:57.0483802Z,
"TaskName": HttpIncomingRequestEndWithClientFailure,
"operationName": GET/SUBSCRIPTIONS/RESOURCEGROUPS/PROVIDERS/MICROSOFT.ANALYSISSERVICES/SERVERS/PROVIDERS/MICROSOFT.INSIGHTS/METRICS,
"apiVersion": 2021-05-01,
"httpMethod": GET,
"httpStatusCode": 400,
"targetUri": [https://management.azure.com:443/subscriptions/subid/resourceGroups/RG/providers/Microsoft.AnalysisServices/servers/servername/providers/providers/Microsoft.Insights/metrics?timespan=2023-03-20%2f2023-03-21&interval=PT1M&metricnames=qpu_metric%2cmemory_metric%2cMemoryLimitHard%2cQueryPoolJobQueueLength&aggregation=Average&api-version=2021-05-01]
"targetResourceProvider": MICROSOFT.INSIGHTS,
"targetResourceType": METRICS,
"userAgent": azsdk-python-azure-mgmt-monitor/6.0.0 Python/3.8.10 (Linux-5.4.0-1100-azure-x86_64-with-glibc2.29)

 
Good one: 
"PreciseTimeStamp": 2023-03-20T04:00:46.8433581Z,
"TaskName": HttpIncomingRequestEndWithSuccess,
"operationName": GET/SUBSCRIPTIONS/RESOURCEGROUPS/PROVIDERS/MICROSOFT.ANALYSISSERVICES/SERVERS/PROVIDERS/MICROSOFT.INSIGHTS/METRICS,
"apiVersion": 2018-01-01,
"httpMethod": GET,
"httpStatusCode": 200,
"targetUri": https://management.azure.com:443/subscriptions/subid/resourceGroups/RG/providers/Microsoft.AnalysisServices/servers/servername/providers/Microsoft.Insights/metrics?timespan=2023-03-19%2f2023-03-20&interval=PT1M&metricnames=qpu_metric%2cmemory_metric%2cMemoryLimitHard%2cQueryPoolJobQueueLength&aggregation=Average&api-version=2018-01-01,
"targetResourceProvider": MICROSOFT.INSIGHTS,
"targetResourceType": METRICS,
"userAgent": azsdk-python-azure-mgmt-monitor/5.0.1 Python/3.8.10 (Linux-5.4.0-1100-azure-x86_64-with-glibc2.29)

No repository file or test is named. Start by comparing the 5.0.1 and 6.0 request URLs and the Azure Monitor metrics API specifications, then reproduce the Databricks query against version 6.0. Done means identifying and correcting the specification or generated request that causes the 400 response, with a regression test showing metric data is returned.

Written by the indexing model from the issue text.