78 строки
3.4 KiB
Markdown
78 строки
3.4 KiB
Markdown
# How to use a VNet
|
|
|
|
If you have a VNET connected to your company network and you need
|
|
to be able to deploy your tests within that VNET, this document will tell
|
|
you how to configure your VNET and RAFT deployment.
|
|
|
|
RAFT assumes that all your test jobs will either be deployed in the VNET or not
|
|
in the VNET. It is a system level configuration, not a job-by-job configuration.
|
|
|
|
Note that deploying into a VNET does take a little longer than deploying outside
|
|
a VNET.
|
|
|
|
#### Step 1: Understanding the limitations
|
|
|
|
RAFT uses Azure Container Instances and they require a specific configuration of your
|
|
VNET with specific limitations. See those limitations [here](https://docs.microsoft.com/en-us/azure/container-instances/container-instances-virtual-network-concepts).
|
|
|
|
When you deploy your RAFT job, it must be deployed into the same region as your VNET.
|
|
This can be accomplished in two different ways.
|
|
|
|
One way is on a per-job basis. Using
|
|
this way, you will use the `--region` parameter whenever you submit your job to specify
|
|
your VNET region.
|
|
|
|
The second way is to deploy your RAFT service into the same region as your VNET. Using this
|
|
method you will not need to use the `--region` parameter as RAFT will default the job's
|
|
region to the system's deployment region. For example if you deploy RAFT to `westus` all
|
|
jobs will also be deployed to `westus` by default.
|
|
|
|
#### Step 2: VNET Configuration
|
|
|
|
Your VNET can exist in any region and in any resource group.
|
|
|
|
**Create a Subnet on your VNET.** When RAFT creates a Container Instance for your job and it is deployed into your VNET and subnet,
|
|
it will be assigned an
|
|
IP address from the subnet address range. Be sure to create a subnet with enough address space to hold
|
|
as many simultaneous jobs as you think you will need. For most deployments 8 bits of
|
|
address space is more than enough.
|
|
|
|
The subnet can be created via the Azure Portal.
|
|
|
|
It is important that the subnet be created only for use with Azure Container Instances.
|
|
You do this by specifing the Subnet Delegation.
|
|
For more information on why, see the limitations referenced above.
|
|
|
|
|
|
|
|
Additionally you must enable the storage service endpoint. RAFT mounts fileshares from
|
|
your deployment's storage account onto the running containers. Setting the service endpoint
|
|
in the subnet enables this behavior.
|
|
|
|
|
|
|
|
#### Step 3: Tell RAFT about your VNET
|
|
|
|
After your VNET is configured and RAFT is deployed, use the following CLI command
|
|
to configure RAFT to use your VNET.
|
|
|
|
`python raft.py service config-vnet --vnetName myVnet --vnetSubnetName myVnetSubnet --vnetResourceGroup myVnetRG --vnetRegion myVnetRegion`
|
|
|
|
| Parameter | Definition |
|
|
|--- | :--|
|
|
| `--vnetName` | The name of your VNET resource in Azure. |
|
|
| `--vnetSubnetName` | The name of the subnet within the VNET. |
|
|
| `--vnetResourceGroup` | The name of the resource group where the VNET lives. |
|
|
| `--vnetRegion` | The region the VNET is deployed. |
|
|
|
|
This command creates a Network Profile that is used when deploying the Container Instance
|
|
and updates the orchestrator's configuration settings `RAFT_NETWORK_PROFILE_NAME` and
|
|
`RAFT_VNET_RESOURCE_GROUP`. These orchestrator settings inform
|
|
the orchestrator how to deploy your job into your VNET.
|
|
|
|
Your VNET will need to have already been configured for this command to succeed.
|
|
|
|
If you decide to remove the VNET configuration from RAFT use the command:
|
|
`python raft.py service clear-vnet --vnetResourceGroup myVnetRG`
|
|
|