Sample - Add a sample for using SAS resource files

samples/sas_resource_files/1989.csv

@@ -0,0 +1,3 @@
+Name,Age
+Julie,16
+John,19

samples/sas_resource_files/1990.csv

@@ -0,0 +1,3 @@
+Name,Age
+Julie,17
+John,20

samples/sas_resource_files/README.md

@@ -0,0 +1,11 @@
+# SAS Resource Files
+
+The following sample shows how to transfer data using secure [SAS blob tokens](https://docs.microsoft.com/en-us/azure/storage/common/storage-dotnet-shared-access-signature-part-1). This allows secure transfer between cloud storage from either your local computer or the nodes in the cluster.
+
+As part of this example you will see how to create a secure write-only SAS and upload files to the cloud. Then create a secure read-only SAS and download those files to the nodes in your cluster. Finally, you will enumerate the files on each node in the cluster and can operate against them however you choose.
+
+Make sure to replace the storage account you want to use. The the storage account listed in the credentials.json file must be used for this sample to work.
+
+```R
+storageAccountName <- "<YOUR_STORAGE_ACCOUNT>"
+```

samples/sas_resource_files/sas_resource_files_cluster.json

@@ -0,0 +1,22 @@
+{
+  "name": "sas_resource_files",
+  "vmSize": "Standard_D11_v2",
+  "maxTasksPerNode": 1,
+  "poolSize": {
+    "dedicatedNodes": {
+      "min": 0,
+      "max": 0
+    },
+    "lowPriorityNodes": {
+      "min": 3,
+      "max": 3
+    },
+    "autoscaleFormula": "QUEUE"
+  },
+  "rPackages": {
+    "cran": [],
+    "github": [],
+    "bioconductor": []
+  },
+  "commandLine": []
+}

samples/sas_resource_files/sas_resources_files_example.R

@@ -0,0 +1,65 @@
+library(doAzureParallel)
+
+doAzureParallel::setCredentials("credentials.json")
+storageAccountName <- "<YOUR_STORAGE_ACCOUNT>"
+inputContainerName <- "datasets"
+
+# Generate a sas tokens with the createSasToken function
+
+# Write-only SAS. Will be used for uploading files to storage.
+writeSasToken <- rAzureBatch::createSasToken(permission = "w", sr = "c", path = inputContainerName)
+
+# Read-only SAS. Will be used for downloading files from storage.
+readSasToken <- rAzureBatch::createSasToken(permission = "r", sr = "c", path = inputContainerName)
+
+# Create a Storage container in the Azure Storage account
+rAzureBatch::createContainer(inputContainerName)
+
+# Upload blobs with a write sasToken
+rAzureBatch::uploadBlob(inputContainerName,
+           fileDirectory = "1989.csv",
+           sasToken = writeSasToken,
+           accountName = storageAccountName)
+
+rAzureBatch::uploadBlob(inputContainerName,
+           fileDirectory = "1990.csv",
+           sasToken = writeSasToken,
+           accountName = storageAccountName)
+
+# Create URL paths with read-only permissions
+csvFileUrl1 <- rAzureBatch::createBlobUrl(storageAccount = storageAccountName,
+              containerName = inputContainerName,
+              sasToken = readSasToken,
+              fileName = "1989.csv")
+
+
+csvFileUrl2 <- rAzureBatch::createBlobUrl(storageAccount = storageAccountName,
+                             containerName = inputContainerName,
+                             sasToken = readSasToken,
+                             fileName = "1990.csv")
+
+# Create a list of files to download to the cluster using read-only permissions
+# Place the files in a directory called 'data'
+resource_files = list(
+  rAzureBatch::createResourceFile(url = csvFileUrl1, fileName = "data/1989.csv"),
+  rAzureBatch::createResourceFile(url = csvFileUrl2, fileName = "data/1990.csv")
+)
+
+# Create the cluster
+cluster <- makeCluster("sas_resource_files_cluster.json", resourceFiles = resource_files)
+registerDoAzureParallel(cluster)
+workers <- getDoParWorkers()
+
+# Files downloaded to the cluster are placed in a specific directory on each node called 'wd'
+# Use the pre-defined environment variable 'AZ_BATCH_NODE_STARTUP_DIR' to find the path to the directory
+listFiles <- foreach(i = 1:workers, .combine='rbind') %dopar% {
+  fileDirectory <- paste0(Sys.getenv("AZ_BATCH_NODE_STARTUP_DIR"), "/wd", "/data")
+  files <- list.files(fileDirectory)
+  df = data.frame("node" = i, "files" = files)
+  return(df)
+}
+
+# List the files downloaded to each node in the cluster
+listFiles
+
+stopCluster(cluster)