Note: The SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics is an example architecture only. It is not intended for production use. Specifically, the Quickstart provides an example of how a SAS Visual Analytics in SAS 9.4 workload and a SAS Visual Statistics in SAS 9.4 workload running in a massively parallel processing (MPP) environment with SAS LASR Analytic Server, can be run on Azure. When running in an MPP environment, the SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics deploys Apache Hadoop. As stated in the SAS 9.4 Support for Hadoop policy, SAS does not provide support for the installation or administration of Apache Hadoop. For support with these issues, please refer to the Apache Hadoop open-source community. The intent of the SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics is to provide customers an opportunity to test deployments on the cloud before committing to a lift and shift from their current on-premises deployments.
- Overview
- Architecture
- Prerequisites
- Deployment Steps
- Post Deployment Steps
- Troubleshooting
- Appendix A: Getting the SAS Token
Note: The SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics is generally available for consideration. Your SAS account team and SAS Enterprise Excellence Center will evaluate if this Quickstart is the recommended solution for you as opposed to the more current SAS Viya Quickstart Template for Azure. If you are interested in this Quickstart, see Contact SAS.
This README for the SAS 9.4M7 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics is used to deploy the following SAS 9.4 products in the Azure cloud:
-
SAS Visual Analytics 7.51
-
SAS Visual Statistics 7.51
This Quickstart is a reference architecture for users who want to deploy SAS Visual Analytics in SAS 9.4 and SAS Visual Statistics in SAS 9.4 using cloud-friendly technologies. By deploying SAS 9.4 in Azure, you get access to SAS analytics visualization and the ability to create powerful statistical models in an environment validated by Azure.
For assistance with SAS software, contact SAS Technical Support. When you contact SAS Technical Support, you will be required to provide information, such as your SAS site number, company name, email address, and phone number, that identifies you as a licensed SAS software customer.
You are responsible for the cost of the Azure services used while running this Quickstart deployment. There is no additional cost for using the Quickstart. You will need a SAS license to launch this Quickstart. Your SAS account team and the SAS Enterprise Excellence Center can advise on the appropriate software licensing and sizing to meet your workload and performance needs as well as if this Quickstart is the recommended solution for you as opposed to the more current SAS Viya Quickstart Template for Azure.
The SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics creates instances as follows:
- one SAS Compute Server virtual machine (VM) for the SAS Visual Analytics in SAS 9.4M7 and SAS Visual Statistics in SAS 9.4M7 server main node, plus the number of VMs for the SAS Visual Analytics and SAS Visual Statistics workers (as specified in the Visual Analytics Worker Count parameter)
- the number of middle-tier VMs for the SAS 9.4M7 mid-tier server (as specified in the Mid-Tier VM Count parameter)
- the number of metadata server VMs for the SAS 9.4M7 metadata clustered server (as specified in the Metadata VM Count parameter)
- one Linux Bastion Host VM
Note: The first release of the SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics supports only a single instance of the metadata server.
To determine the appropriate sizes of these resources, we recommend that you work with your SAS account team to form a hardware estimate and use the instance sizes from the hardware estimate. See Best Practices for Using Microsoft Azure with SAS for more information.
Note: While the Standard_B2ms VM size is available, it should only be used to create a deployment without a SAS installation (i.e. leave the SAS Depot Location parameter blank) for exploratory reasons. Using the Standard_B2ms size with a SAS installation will result in a failed deployment, as that size is too small to support SAS.
This SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics takes a license for SAS 9.4 and deploys SAS into its own network. The deployment creates the network and other infrastructure. After the deployment process completes, you will have the outputs for the web endpoints for a SAS 9.4 deployment on recommended VMs.
For details, see SAS 9.4 Intelligence Platform: Installation and Configuration Guide.
By default, Quickstart deployments enable Transport Layer Security (TLS) to help ensure that communication between external clients (on the internet) and the load balancer is secure. Likewise, TLS is enabled between the load balancer and the private subnet that contains the SAS 9.4 components.
Deploying this Quickstart for a new virtual private cloud (VPC) with default parameters in a massively parallel processing (MPP) environment builds the following SAS 9.4 environment in the Microsoft Azure cloud, shown in Figure 1.
Note: In the diagram below, a SAS 9.4M7 metadata clustered server is depicted. However, the first release of the SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics supports only a single instance of the metadata server.
Figure 1: Quickstart Architecture in Azure for SAS Visual Analytics in SAS 9.4 and SAS Visual Statistics in SAS 9.4 in an MPP Environment
Before deploying the SAS 9.4 Quickstart Template for Azure with SAS Visual Analytics and SAS Visual Statistics, you must have the following:
-
A Microsoft Azure account with Contributor and Admin roles if you do not already have one.
-
A SAS Visual Analytics in SAS 9.4M7 and SAS Visual Statistics in SAS 9.4M7 software order that contains the following supported Quickstart products:
SAS Visual Analytics 7.51 on Linux SAS Visual Statistics 7.51 on Linux -
A software depot that has been downloaded with the SAS Download Manager (per the instructions in the Software Order Email) to a folder similar to this: /<download_folder_name>/nondist/depot.
-
Sufficient number of cores. We recommend that you work with your SAS account team to form a hardware estimate, and use the instance sizes from the hardware estimate. See Best Practices for Using Microsoft Azure with SAS for more information.
-
Verification that your required SAS 9.4M7 file upload sizes do not exceed the limits of the Application Gateway. For details about limits, see "Application Gateway limits."
-
A resource group that does not already contain a Quickstart deployment. For more information, see "Resource groups".
-
Because this Quickstart uses Azure NFS file shares, you must enable your Azure subscription for NFS. For more information, see the "Register the NFS 4.1 protocol" section in How to create an NFS File Share.
-
Use of a region that supports Availability Zones. For more information, see Azure Services that support Availability Zones.
-
The software depot must be uploaded to Azure Blob Storage as follows:
- Navigate to the directory where the software depot was downloaded.
- Upload the software depot by running this command:
az storage blob upload-batch --account-name "$STORAGE_ACCOUNT" --account-key "$STORAGEKEY" --destination "$SHARE_NAME" --destination-path "$SUBDIRECTORY_NAME" --source "$(pwd)"
Here is a description of the variables in the above command:
STORAGE_ACCOUNT - the name of the Azure storage account to which the depot is uploaded.
STORAGEKEY - the access key for the Azure storage account. This value can be retrieved using the Azure CLI az storage account keys list command. See az storage account keys for more information.
SHARE_NAME - the name of the Azure container to which the depot is uploaded. This container must exist in the storage account.
SUBDIRECTORY_NAME - the name of the Azure storage blob to which the depot is uploaded. This storage blob is created by the az storage account keys list command.
For more information about this command, see az storage blob upload-batch.
- You must obtain the URL for the SAS depot location as follows:
Note: In these steps, SAS Token refers to the Shared Access Signature Token.
- From the Azure Portal here, click Resource groups.
- From the left column, click on the storage account that you uploaded the SAS 9.4 order files to.
- Expand BLOB CONTAINERS. Verify that your order files are inside the blob container.
- Save the SAS Token string. See Appendix A for instructions.
- Click Copy URL from the menu at the top to get the path to your order files.
- Append the SAS Token string that you saved in step 4 to the end of the URL that you copied in step 5 to determine the URL for the SAS software depot location.
For example, suppose that the URL path to your order files is: <protocol>://sas9storageaccount.blob.core.windows.net/sas9-storagecontainer/<order-directory>
Suppose that the SAS Token string is <token-string>.
Then the SAS software depot location is: <protocol>://sas9storageaccount.blob.core.windows.net/sas9-storagecontainer/<order-directory><token-string>
Note: We recommend that you run only one installation of the Quickstart at a time.
- Click the Deploy to Azure button at the beginning of this document here.
- Specify the following parameters for your deployment:
| Parameter Name | Value |
|---|---|
| Subscription | Specifies what subscription to use for the deployment. |
| Resource group | Specifies what resource group to use. Choose an existing group or click Create new and provide a name for the new group. |
| Region | Defines the Azure region in which the deployment should run. The available Azure regions are the ones listed at both Azure Services that support Availability Zones, and How to create an NFS share. |
| Location | Defines the location in Azure where these resources should be created. This is derived from the resource group. |
| SAS Depot Location | Specifies the URI of the Azure Blob Store where the software depot was uploaded. You determined this URI during the prerequisite step here. |
| SAS Plan File Location | Specifies the URI to download the plan file from as currently specified in the Azure Blob Store. Leave this blank if the plan file is stored in the software depot blob. If this is left blank, then the plan file located in the depot at depot/plan.xml is used. If no plan file exists in the depot, then the default plan file supplied by this Quickstart is used. |
| Use a New or Existing Virtual Network | Specifies whether to use a new or existing network. |
| Existing Virtual Network Resource Group | Specifies the resource group if using an existing virtual network. Leave this blank if you are using a new network. Otherwise, enter the resource group for the existing network |
| Virtual Network Name | Use the default value (recommended). |
| Operating System Image | Specifies the operating system to use. Currently, only SUSE is supported. |
| Mid-Tier VM Count | Specifies the number of virtual machines (VMs) for the middle tier server. Select 1 for a non-clustered middle tier server. Select 2 for a two node middle tier cluster. |
| Mid-Tier VM Size | Specifies the VM size. Use the default size (recommended). |
| Visual Analytics Worker Count | Specifies the number of worker instances created for the SAS Visual Analytics controller. |
| Visual Analytics Controller Size | Specifies the size of the Visual Analytics Controller. |
| Visual Analytics Worker Size | Specifies the size of the Visual Analytics Worker. |
| Proximity Placement Group Name | Specifies the proximity group for instances. You provide the name. |
| SSH Key for VM Access | Specifies the full SSH public key that will be added to the servers. Copy and paste a public SSH key into this field. |
| SAS Administration Password | Specifies the password used for SAS authentication. Enter the password to be used for the sasadm@saspw account. |
| Azure Administration Password | Specifies the password used for OS authentication. Enter the password to be used for the sasinst account. |
| Admin Ingress Location | Specifies to allow inbound SSH traffic to the Ansible Controller from this Classless Inter-Domain Routing (CIDR) block (IP address range). Must be a valid IP CIDR range of the form x.x.x.x/x. If this is left blank, the environment can be accessed from any location on the internet. |
| Web Ingress Location | Specifies to allow inbound HTTP traffic to the SAS 9.4 environment from this CIDR block (IP address range). Must be a valid IP CIDR range of the form x.x.x.x/x. If this is left blank, the environment can be accessed from any location on the internet. |
| _artifacts Location SAS Token | Leave this blank. Note: SAS Token refers to the Shared Access Signature Token |
| _artifacts Location | Use the default value (recommended). |
- Click Next: Review and Create.
- If the validation is successful, click Create. When the Deployment is in progress window appears, the deployment begins. Deployments typically take two to three hours to complete.
After deploying, an Instructions.html file is created in the /opt/sas/config/Lev1/Documents directory on each SAS VM. You are directed to follow the instructions in this file as a post-installation step. However, the host names referenced in the Instructions.html file URLs are not accessible from external clients. Instead, construct the URLs using the base URLs referenced in the steps below.
- Open the resource group.
- Click Deployments.
- Click Microsoft Template-<deployment name>.
- Click Outputs.
- From the list of URLs that appear, copy the base URL from the sasVisualAnalyticsHub or sasStudio URLs. Substitute that value for the base URLs in the Instructions.html file. In addition, make the following changes to the base URLs in the Instructions.html file:
- If the URL path in the Instructions.html file has port 7980, remove the port as it is not valid in the Quickstart.
- If the URL path in the Instructions.html file has port 7080, replace with port 8441.
- Change the protocol in all URLs from HTTP to HTTPS.
For example, suppose that the URL path to the SAS Visual Analytics Hub in the Outputs window is: https://sas94-dpm1234567890-123e3.eastus.cloudapp.azure.com/SASVisualAnalyticsHub
The base URL portion of the URL above is: https://sas94-dpm1234567890-123e3.eastus.cloudapp.azure.com
Suppose that the URL path to SAS Visual Data Builder in the Instructions.html file is: https://midtier-0.internal.cloudapp.net:7980/SASVisualDataBuilder
The base URL portion of the URL above is: https://midtier-0.internal.cloudapp.net
Substitute the base URL from the Outputs window for the base URL in the Instructions.html file: https://sas94-dpm1234567890-123e3.eastus.cloudapp.azure.com/SASVisualDataBuilder
Note: The port (:7980) in the URL from the Instructions.html is not included in the new URL.
- Open the resource group.
- Click Deployments.
- Click Microsoft Template-<deployment name>.
- Click Outputs to access the following:
- jump_IP IP address
- SAS Visual Analytics URL
- SAS Studio URL
- SAS Environment Manager
- Navigate to the SAS Visual Analytics, SAS Studio, and SAS Environment Manager URLs. Log in as sasadm@saspw. Enter the password that you specified for the SAS Administration Password value here.
The easiest way to run SAS Management Console is to connect to the deployment using X11 port forwarding as follows:
ssh -X -i <public key pem file> AzureUser@<jumpvm public IP address>
ssh -X <vm name>
cd /opt/sas/SASHome/SASManagementConsole/9.4
./sasmc &
For example, to run SAS Management Console on the midtier-0 VM:
ssh -X -i <public key pem file> AzureUser@<jumpvm public IP address>
ssh -X midtier-0
cd /opt/sas/SASHome/SASManagementConsole/9.4
./sasmc &
Check the logs and services status by accessing the various VM instances from the jumpvm, as follows:
-
Obtain the public IP address of the jumpvm from jumpvm details:
a. Navigate to https://portal.azure.com/#blade/HubsExtension/BrowseResourceGroups and select the resource group created by the deployment.
b. Select the jumpvm resource from the list of resources in resource groups. The public IP address is displayed in the Overview.
-
Using the public IP address, SSH to the jumpvm:
ssh -i <public key pem file> AzureUser@<jumpvm public IP address>The jumpvm provides SSH access to the other VMs in the deployment:
- SSH access using the AzureUser account
ssh <vm name>For example:
ssh midtier-0- SSH access using the sasinst service account
ssh sasinst@<vm name>For example:
ssh sasinst@midtier-0The password for all system accounts is set to the value provided by the user in the Azure Administration Password parameter.
Some services might not successfully restart when the deployment completes.
To restart the services using the start and stop scripts, perform the following steps:
-
Obtain the public IP Address of the jumpvm from the jumpvm’s details:
a. Navigate to https://portal.azure.com/#blade/HubsExtension/BrowseResourceGroups and select the resource group created by the deployment.
b. Select the jumpvm resource from the list of resources in the Resource Group. The public IP address is displayed in the Overview.
-
Using the public IP address, SSH to the jumpvm:
ssh -i <public key pem file> AzureUser@<jumpvm public IP address> -
Log in to the Azure CLI:
/usr/local/bin/az login --use-device-code
For more information about the Azure CLI, see Sign in with Azure CLI.
- Verify that the current subscription matches the subscription in which SAS has been deployed:
/usr/local/bin/az/ account list --output table
For more information about the az account command see az account.
- If the subscriptions do not match, set the current subscription to the subscription in which SAS has been deployed:
/usr/local/bin/az/ account set --subscription <subscription>
-
When the correct subscription is set, you can run the start and stop scripts as follows:
a. To start the SAS services on the SAS 9.4M7 Metadata Server, SAS 9.4M7 Mid-Tier Server, and SAS Visual Analytics controller VMs:
/sas/install/scripts/start_sas.shb. To return the status of the SAS services on the SAS 9.4M7 Metadata Server, SAS 9.4M7 Mid-Tier Server, and SAS Visual Analytic controller VMs:
/sas/install/scripts/status_sas.shc. To stop the SAS services on the SAS 9.4M7 Metadata Server, SAS 9.4M7 Mid-Tier erver, and SAS Visual Analytic controller VMs:
/sas/install/scripts/stop_sas.sh
Note: In these steps, SAS Token refers to the Shared Access Signature Token.
- From the Azure Portal here, click Resource groups.
- From the left column, click on the storage account that you uploaded the SAS 9.4 order files to.
- Expand BLOB CONTAINERS. Verify that your order files are inside the blob container.
- From the left column, click on Shared access signature.
- From the Shared access signature window, select the Container and Object check boxes under Allowed resource types.
- Specify an end date. The end date specifies the expiration date of your SAS Token. You might want to specify a date that is a year in the future.
- Copy and paste the string under SAS Token. Save to a known location.