Microsoft Azure® is a commercial cloud computing platform upon which the Liquids HCA Tool performs Land Spill GeoClaw overland flow simulations for your release points.
To run GeoClaw simulations on Azure you must have an Azure portal subscription with access to an Azure Batch account properly configured to run GeoClaw, together with an associated Azure Storage account. This help topic walks you through the steps of connecting to the Azure portal and configuring a new Azure Batch account resource for use with the Liquids HCA Tool. It is assumed that you have an existing Azure portal subscription, that your user credentials are linked to this subscription, and that your user privileges are sufficient to allow the creation of new Azure resources. For illustration purposes, we’ll assume that you work for the ACME company and that your Microsoft account is email@example.com.
To learn more about creating an Azure Batch account, see the Azure help topic, Create a Batch account with the Azure portal.
Connect to the Azure portal
You interact with Azure cloud services via a web-based Graphical User Interface (GUI). To connect to the Azure portal, enter the following URL address in your browser: http://portal.azure.com. Unless you are already connected, you will be redirected to the Microsoft login page:
Select or enter your account. Once you’ve selected your account, you’ll be prompted to enter your password. Upon entering your password, press the ‘Sign in’ button to enter the Azure Portal:
Create an Azure Batch account
On the Azure portal homepage, select the ‘Create a resource’ button in the ‘Azure Services’ section of the window:
Selecting the ‘Create a resource’ button opens the Azure portal ‘New’ resource window. In the ‘Search the Marketplace’ text entry box, type ‘Batch Service’ and select the ‘Batch Service’ resource from the search results list:
Selecting the ‘Batch Service’ resource takes you to the ‘Batch Service’ resource creation page:
Select the ‘Create’ button to begin the instantiation of your new Azure Batch Service via the ‘New Batch account’ window:
You’ll need to set the following properties for your new Azure Batch account:
- Subscription – Your organization may have more than one Azure subscription. Make sure to select the right subscription! In this case, we’re going to use ACME’s DevOps subscription.
- Resource group – Assigning your new Batch account to a resource group is optional, but it’s highly recommended if yours is a large organization with a large number of Azure resources. In this example, we’ve created a ‘LiquidsHCA’ resource group under which to organize Liquids HCA Azure resources.
- Account name – Azure Batch account names must be unique; note that the account name is concatenated with the Azure region. Azure will check your name for uniqueness and warn you if you need to modify it. The Batch account name is not important to processing; you can name it anything you want. In our example, we’ve named the Batch account ‘landspill.’
- Location – You must specify an Azure region in which to locate the Batch service. You can locate the Batch account in any Azure region, but it’s generally a good idea to pick the region in which you are located, to optimize GeoClaw results download speed. Here, we’ve selected the ‘Central US’ region.
Your new Liquids HCA Azure Batch account must have an associated Azure Storage account, and it should be in the same region as the Batch account (to optimize data transfer from your Batch account compute pool nodes to storage). You can associate any Azure Storage account with your new Azure Batch account; but, for our purposes, we’ll assume that you are creating a new, dedicated Azure Storage account for use with your new Azure Batch account. Once you’ve set your basic Azure Batch account properties, click on the ‘Select a storage account’ hyperlink to open the ‘Choose storage account’ window:
Select the ‘Create new’ button to open the ‘Create storage account’ window:
As with your new Azure Batch account, the name of your new Storage account must be unique. As with your Batch account name, Azure will check your Storage account name for uniqueness and warn you if you need to modify it. The default values for the remaining Storage account properties are sufficient to the needs of the Liquids HCA Tool. Select the ‘OK’ button to create the new Azure Storage account and return to the ‘New Batch account’ window. Note that your newly created Azure Storage account is now associated with your Azure Batch account:
Select the ‘Next: Advanced >’ button to open the ‘Advanced’ section of the ‘New Batch account’ window:
You should use the default settings unless your organization has specific reasons to alter them. Note that the Liquids HCA Tool requires that ‘Pool allocation mode’ be set to ‘Batch service.’ Select the ‘Next: Tags >’ button to advance to the ‘Tags’ section of the ‘New Batch account’ window:
Tags are an optional way of grouping resources; they give you a way of mixing and matching resources across and between resource groups, primarily to give you a mechanism to slice and dice data for tracking and reporting purposes. In this simple example, tags are not necessary and in no way affect GeoClaw processing, so we’ll not add any tags to your new Azure Batch account. Select the ‘Next: Review + create >’ button to move on to the ‘Review + create’ section of the ‘New Batch account’ window:
Select the ‘Create’ button to create your new Azure Batch account!
It will take Azure a few minutes to deploy your new Batch account. When the deployment is complete, you should see a screen similar to this:
Using your Azure Batch account with the Liquids HCA Tool
After you’ve created your Azure Batch account, you may wish to provide access to it for additional users. You can do this by selecting the ‘Access control (IAM)’ item from your Batch account properties items:
Selecting the ‘Access control (IAM)’ property item for your Azure Batch account opens the ‘Access control (IAM)’ window for your Batch account:
You can use this window to provide additional users with controlled access to your Azure Batch account.
The Liquids HCA Tool accesses your Azure Batch account and associated Storage account via your Batch and Storage accounts’ Azure credentials. You can view your Azure Batch and Storage account credentials by selecting the ‘Keys’ item from your Batch account properties items:
Selecting the ‘Keys’ property item for your Azure Batch account opens the ‘Keys’ window for your Batch account:
If you choose to do so, you can store your Azure Batch and Storage account credentials in an encrypted, passcode-protected file that you create using the Create Encrypted Azure Credential File tool. The encrypted, passcode-protected Azure credentials file simplifies the process of passing your Azure Batch and Storage account credentials to the Liquids HCA Tools geoprocessing tools that use them.
When GeoClaw executes on your Azure Batch account, it automatically creates a computing pool called landspill-azure-pool. The properties of the pool are set using Create GeoClaw Cases. The pool is instantiated by the Run Cases on Azure tool. You can view the characteristics of the landspill-azure-pool Azure pool by selecting the ‘Pools’ item from your Batch account properties items. Run Cases on Azure also creates the landspill-azure-job Azure job under your Azure Batch account; each of your GeoClaw cases becomes a task under the landspill-azure-job Azure job. You can check on the status of your running landspill-azure-job Azure job by selecting the ‘Jobs’ item from your Batch account properties items.
The number of nodes available in the landspill-azure-pool Azure pool is determined by the quota limits of your Azure Batch account. You can see how your Azure Batch account quotas are assigned by selecting the ‘Quotas’ item from your Batch account properties items:
Selecting the ‘Quotas’ property item for your Azure Batch account opens the ‘Quotas’ window for your Batch account:
You can view your Azure Batch account’s current virtual CPU (vCPU) quotas and their distribution among the various Azure Virtual Machine (VM) types in the ‘Quotas’ window for your Batch account. Note that you’ll want to change your Batch account quota settings immediately to make sure the correct VM type (H Series) for the Liquids HCA Tool is available for use in your Azure Batch account.
Request a quota change for your Azure Batch account
The default number of vCPUs on your Azure Batch account may vary depending on your billing mechanism. By default, your available Batch account vCPUs are all devoted to the Dv3 Series VM type. The Liquids HCA Tool runs best on H Series VMs, so you should reallocate all your available vCPUs to the H Series VM type. Be aware that some quota changes/increases may require a change in your billing mechanism. You can request a quota change and/or increase by clicking on the ‘Request Quota Increase’ button in the ‘Quotas’ window for your Batch account:
Select the ‘Request quota increase’ button to open the ‘New support request’ window:
The ‘Subscription’ value should be automatically set to the subscription under which your Azure Batch account falls. Enter the following values for the remaining parameters:
- Issue type – Service and subscription limits (quotas)
- Quota type – Batch
Select the ‘Next: Solutions >>’ button to move on to the ‘Details’ section of the ‘New support request’ window:
Set the ‘SUPPORT METHOD’ parameters as desired and then click on the ‘Enter details’ hyperlink to open the ‘Quota Details’ panel of the ‘New support request’ window:
You’ll need to select the ‘Location’ (region) for your Azure Batch account. Our example ‘landspill’ Batch account is in the ‘Central US’ region. The ‘Quota type’ value should be ‘Per Batch account.’ You need to allocate all your vCPUs to the H Series VM type, and you also want to increase your vCPU quota to 40, since the preferred H8 VM has eight vCPUs. To initialize the change, check the Dv3 Series and H Series checkboxes in the ‘Select Quotas to Update’ pull down to indicate the items to which you want to make changes. Select the ‘Save and continue’ button to reduce the list to only those items you wish to change:
In this example, you decrease the Dv3 Series vCPU count to zero (0) and increase the H Series vCPU count to forty (40). Select the ‘Save and continue’ button to return to the ‘Details’ section of the ‘New support request’ window:
Note that your request quota changes are recorded in the ‘Details’ section of the ‘New support request’ window. Press the ‘Nest: Review + create >>’ button to open the ‘Review + create’ section of the ‘New support request’ window:
Review your quota change request details to make sure everything is correct, and select the ‘Create’ button to submit your request. Your Azure Batch account quota change request will be managed as an Azure support request.