public static async Task Run(
[EventHubTrigger("%EventHubName%", Connection = "EventHubConnectionString", ConsumerGroup = "%EventHubConsumerGroup%")] string eventHubMessages,
ILogger log)
{
try
{
log.LogInformation("CollectMetrics function started.");
IoTHubMetric[] iotHubMetrics = JsonConvert.DeserializeObject <IoTHubMetric[]>(eventHubMessages);
IEnumerable <LaMetric> metricsToUpload = iotHubMetrics.Select(m => new LaMetric(m, string.Empty));
LaMetricList metricList = new LaMetricList(metricsToUpload);
// initialize log analytics class
AzureLogAnalytics logAnalytics = new AzureLogAnalytics(
workspaceId: _workspaceId,
workspaceKey: _workspaceKey,
logger: log,
apiVersion: _workspaceApiVersion);
bool success = false;
for (int i = 0; i < Constants.UploadMaxRetries && (!success); i++)
{
// TODO: split up metricList so that no individual post is greater than 1mb
success = await logAnalytics.PostToInsightsMetricsAsync(JsonConvert.SerializeObject(metricList), _hubResourceId, _compressForUpload);
}
if (success)
{
log.LogInformation($"Successfully sent {metricList.DataItems.Count()} metrics to fixed set table");
}
else
{
log.LogError($"Failed to send {metricList.DataItems.Count()} metrics to fixed set table after {Constants.UploadMaxRetries} retries");
}
}
catch (Exception e)
{
log.LogError($"CollectMetrics failed with the following exception: {e}");
}
}
public static async Task Run(
[QueueTrigger("%QueueName%", Connection = "StorageConnectionString")] string queueItem,
ILogger log)
{
try
{
#region Queue trigger
JObject storageEvent = JsonConvert.DeserializeObject <JObject>(queueItem);
var match = Regex.Match(storageEvent["subject"].ToString(), "/blobServices/default/containers/(.*)/blobs/(.*)", RegexOptions.IgnoreCase);
if (!match.Success)
{
log.LogWarning($"Unable to parse blob Url from {storageEvent["subject"]}");
return;
}
if (!string.Equals(match.Groups[1].Value, _containerName))
{
log.LogDebug($"Ignoring queue item because it is related to container '{match.Groups[1].Value}'");
return;
}
string blobName = match.Groups[2].Value;
#endregion
log.LogInformation($"ProcessModuleLogs function received a new queue message from blob {blobName}");
// Create a BlobServiceClient object which will be used to create a container client
BlobServiceClient blobServiceClient = new BlobServiceClient(_connectionString);
// Create container client object
BlobContainerClient containerClient = blobServiceClient.GetBlobContainerClient(_containerName);
// Get blob client object
BlobClient blobClient = containerClient.GetBlobClient(blobName);
// Read the blob's contents
Stream blobStream = await blobClient.OpenReadAsync();
// Decompress if encoding is gzip
if (string.Equals(_logsEncoding, "gzip", StringComparison.OrdinalIgnoreCase))
{
// TODO: decompress logs
}
StreamReader reader = new StreamReader(blobStream);
IoTEdgeLog[] iotEdgeLogs = JsonConvert.DeserializeObject <IoTEdgeLog[]>(reader.ReadToEnd());
// Convert to logs their final log analytics format
LogAnalyticsLog[] logAnalyticsLogs = iotEdgeLogs.Select(x => new LogAnalyticsLog(x)).ToArray();
if (logAnalyticsLogs.Length == 0)
{
return;
}
// initialize log analytics class
AzureLogAnalytics logAnalytics = new AzureLogAnalytics(
workspaceId: _workspaceId,
workspaceKey: _workspaceKey,
logger: log,
apiVersion: _workspaceApiVersion);
// because log analytics supports messages up to 30MB,
// we have to break logs in chunks to fit in on each request
byte[] logBytes = Encoding.UTF8.GetBytes(JsonConvert.SerializeObject(logAnalyticsLogs));
double chunks = Math.Ceiling(logBytes.Length / (_logMaxSizeMB * 1024f * 1024f));
// get right number of items for the logs array
int steps = Convert.ToInt32(Math.Ceiling(logAnalyticsLogs.Length / chunks));
int count = 0;
do
{
int limit = count + steps < logAnalyticsLogs.Length ? count + steps : logAnalyticsLogs.Length;
log.LogInformation($"Submitting data collection request for logs {count + 1} - {limit} / {logAnalyticsLogs.Length}");
LogAnalyticsLog[] logsChunk = logAnalyticsLogs.Skip(count).Take(limit).ToArray();
try
{
//logAnalytics.Post(JsonConvert.SerializeObject(logsChunk), _logType, _hubResourceId);
bool success = logAnalytics.PostToCustomTable(JsonConvert.SerializeObject(logsChunk), _logType, _hubResourceId);
if (success)
{
log.LogInformation("ProcessModuleLogs request to log analytics completed successfully");
}
else
{
log.LogError("ProcessModuleLogs request to log analytics failed");
}
}
catch (Exception e)
{
log.LogError($"ProcessModuleLogs failed with exception {e}");
}
count += steps;
}while (count < iotEdgeLogs.Length);
// Delete blob after being processed
await blobClient.DeleteIfExistsAsync(Azure.Storage.Blobs.Models.DeleteSnapshotsOption.IncludeSnapshots);
}
catch (Exception e)
{
log.LogError($"ProcessModuleLogs failed with the following exception: {e}");
}
}
