Skip to content

Latest commit

 

History

History
238 lines (158 loc) · 12.9 KB

ResourceGraphDriver.rst

File metadata and controls

238 lines (158 loc) · 12.9 KB
Raw

Azure Resource Graph Provider

This data provider allows for a connection to the Azure Resource Graph and a way to query against the Azure Resource Graph Explorer. The data connector functions in the same way as other data connectors and uses the Kusto Query Language (KQL) and has with some subtle differences to other connectors in they way that authentication is handled.

You would use this data connector to flexibly and quickly get details on deployed Azure resources within a subscription. It allows for bulk queries on various aspects of resources and returns data in a very structured format. This makes it much more effective and efficient than getting resource specific details via the resource API.

More details about data providers in MSTICPy can be found in the documentation <DataProviders>.

Azure Resource Graph Configuration

The provider uses the Azure configuration specified in the Azure section of the msticpyconfig.yaml.

These settings affect all Azure components.

For more information on using and configuring msticpyconfig.yaml see msticpy Package Configuration <../getting_started/msticpyconfig> and MSTICPy Settings Editor<../getting_started/SettingsEditor>

An example.

Azure:
  auth_methods:
  - cli
  - interactive
  cloud: global

For more details see ../getting_started/AzureAuthentication and getting_started/SettingsEditor:Azure Cloud and Authentication Settings.

Loading a Query Provider for Azure Resource Graph

The provider for the Azure Resource Graph is named ResourceGraph

from msticpy.data.data_providers import QueryProvider
qry_prov = QueryProvider("ResourceGraph")

Connecting to Azure Resource Graph

Once initialized the first step in using the data provider is to authenticate. The Resource Graph provider uses MSTICPy's Azure authentication features <data_acquisition/AzureData:Instantiating and Connecting with an Azure Data Connector> and you can provide a set of authentication methods when connecting. By default the provider will attempt to authenticate using credentials stored in msticpyconfig.yaml (or as environment variables) and an Azure CLI connection but this can be customized with the 'auth_methods' keyword.

Once successfully connected you will be presented with a “Connected” message.

qry_prov.connect(auth_methods=["cli"])

Connected

Listing available queries

As with other data providers there are a number of built-in queries with this provider. Once connected you can view the available queries with QUERY_PROVIDER.list_queries().

Alternatively you can view query details in an interactive widget with QUERY_PROVIDER.browse_queries()

For more information, see: data_acquisition/DataProviders:Listing Available Queries.

qry_prov.list_queries()
['ResourceGraph.list_detailed_virtual_machines',

'ResourceGraph.list_public_ips', 'ResourceGraph.list_resources', 'ResourceGraph.list_resources_by_api_version', 'ResourceGraph.list_resources_by_type', 'ResourceGraph.list_virtual_machines']

Running pre-defined query

In order to run pre-defined query, execute with the query name, e.g. QUERY_PROVIDER.ResourceGraph.QUERY_NAME(). You can pass parameters to these queries to customize them, however they will also run with default parameters if none as provider. The query browser will provide details as to what parameters are available with each query.

As with other data providers data is returned to you in a Pandas DataFrame.

For more information , refer documentation - Running a pre-defined query <data_acquisition/DataProviders:Running a pre-defined query>

qry_prov.ResourceGraph.list_resources_by_api_version()
type apiVersion
microsoft.alertsmanagement/actionrules 2019-05-05-preview
microsoft.alertsmanagement/smartdetectoralertrules 2021-04-01
microsoft.apimanagement/service 2019-12-01
microsoft.automanage/accounts 2020-06-30-preview
microsoft.automation/automationaccounts 2018-06-30

Running an ad-hoc query

You can also define a your own KQL query for the Resource Graph and run with QUERY_PROVIDER.exec_query(QUERY)

query = "Resources | where type =~ 'Microsoft.Compute/virtualMachines' | take 3"
qry_prov.exec_query(query)
id name type tenantId kind location resourceGroup subscriptionId managedBy sku plan zones

/subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/LinuxTestLab/providers/Microsoft.Compute/virtualMachines/RHEL77Base /subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/LinuxTestLab/providers/Microsoft.Compute/virtualMachines/Ubuntu18ASC /subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/monster-island/providers/Microsoft.Compute/virtualMachines/GodzillaTron1

RHEL77Base Ubuntu18ASC GodzillaTron1

microsoft.compute/virtualmachines microsoft.compute/virtualmachines microsoft.compute/virtualmachines

1d228542-43d3-43fa-b9f6-e2a5e3d69d47 1d228542-43d3-43fa-b9f6-e2a5e3d69d47 1d228542-43d3-43fa-b9f6-e2a5e3d69d47

eastus2 eastus2 japanwest

linuxtestlab linuxtestlab monster-island

1d228542-43d3-43fa-b9f6-e2a5e3d69d47 1d228542-43d3-43fa-b9f6-e2a5e3d69d47 1d228542-43d3-43fa-b9f6-e2a5e3d69d47

Note: Resource Graph queries are limited to 1000 rows of output each. If your query returns 1000 rows it is likely it has hit this limit, consider re-writing the query to return a smaller subset of data. This applies to both built in queries and ad-hoc queries.

End-to-end Example

In this example we want to take a look at all of the virtual machines we have in our environment and they get specific details including public IP on one of them:

from msticpy.data.data_providers import QueryProvider
# Initialize and connect to provider
qry_prov = QueryProvider("ResourceGraph")
qry_prov.connect()

Connected

# Get list of VMs and see how many we have
vms = qry_prov.ResourceGraph.list_virtual_machines()
print(f"Number of VMs found : {len(vms.index)}")
# Filter the query to get a smaller dataset
vms = qry_prov.ResourceGraph.list_virtual_machines(add_query_items="| where resourceGroup contains 'msticpy'")
display(vms)
# Set hostname for our next query
hostname = vms.iloc[0]['name']

Number of VMs found : 421
id name type tenantId kind location resourceGroup subscriptionId managedBy sku plan zones
/subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/MSTICpy/providers/Microsoft.Compute/virtualMachines/MSTIC-DSVM MSTIC-DSVM microsoft.compute/virtualmachines 1d228542-43d3-43fa-b9f6-e2a5e3d69d47 eastus msticpy 1d228542-43d3-43fa-b9f6-e2a5e3d69d47 ['1']

Now we can get details on the specific VM using its hostname.

qry_prov.ResourceGraph.list_detailed_virtual_machines(host_name=hostname)
vmId vmName vmSize nicId publicIpId publicIpAddress
/subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/MSTICpy/providers/Microsoft.Compute/virtualMachines/MSTIC-DSVM MSTIC-DSVM Standard_B2s /subscriptions/1d228542-43d3-43fa-b9f6-e2a5e3d69d47/resourceGroups/MSTICpy/providers/Microsoft.Network/networkInterfaces/mstic-dsvm832 /subscriptions/40dcc8bf-0478-4f3b-b275-ed0a94f2c013/resourceGroups/MSTICpy/providers/Microsoft.Network/publicIPAddresses/MSTIC-DSVM-ip 11.11.11.111

Other Azure Resource Graph Documentation

Built-in data_acquisition/DataQueries:Queries for Azure Resource Graph.

:pyAzure Resource Graph driver API documentation<msticpy.data.drivers.resource_graph_driver>