Azure Table Storage: track retries - azure

Question

I'm using the standard example from microsoft for inserting new entitys to a table.
Is there a way to track if a retry was performed?
Code:
CloudTableClient tableClient = storageAccount.CreateCloudTableClient();
CloudTable table = tableClient.GetTableReference("people");
CustomerEntity customer1 = new CustomerEntity("Harp", "Walter");
TableOperation insertOperation = TableOperation.Insert(customer1);
table.Execute(insertOperation);
With TransientFaultHandlingFramework it was easy to do:
var retryPol = new RetryPolicy<SqlAzureTransientErrorDetectionStrategy>(retryStrategy);
retryPol.Retrying += (obj, eventArgs) =>
{
var msg = String.Format("Retrying, CurrentRetryCount = {0} , Delay = {1}, Exception = {2}", eventArgs.CurrentRetryCount, eventArgs.Delay, eventArgs.LastException.Message);
System.Diagnostics.Debug.WriteLine(msg);
};

Answer 1

You can use the Enterprise Library Transient Fault Handling Application Block as explained in Rob's answer:
var retryPol = new RetryPolicy<StorageTransientErrorDetectionStrategy>(retryStrategy);
retryPol.Retrying += (obj, eventArgs) =>
{
var msg = String.Format("Retrying, CurrentRetryCount = {0} , Delay = {1}, Exception = {2}", eventArgs.CurrentRetryCount, eventArgs.Delay, eventArgs.LastException.Message);
System.Diagnostics.Debug.WriteLine(msg);
};
var options = new TableRequestOptions { RetryPolicy = new RetryPolicies.NoRetry() };
CloudTableClient tableClient = storageAccount.CreateCloudTableClient();
CloudTable table = tableClient.GetTableReference("people");
CustomerEntity customer1 = new CustomerEntity("Harp", "Walter");
TableOperation insertOperation = TableOperation.Insert(customer1);
retryPol.ExecuteAction<TableResult>(() => { return table.Execute(insertOperation, options); });
If you prefer to use exclusively the Windows Azure Storage Client Library, you can create a custom retry policy that raises events, like this:
public class EventExponentialRetry : IRetryPolicy
{
private static readonly TimeSpan DefaultClientBackoff = TimeSpan.FromSeconds(4.0);
private const int DefaultClientRetryCount = 3;
private TimeSpan deltaBackoff;
private int maximumAttempts;
private ExponentialRetry retry;
public event EventHandler<RetryEventArgs> RaiseRetryEvent;
public EventExponentialRetry()
{
Initialize(DefaultClientBackoff, DefaultClientRetryCount);
}
public EventExponentialRetry(TimeSpan deltaBackoff, int maxAttempts)
{
Initialize(deltaBackoff, maxAttempts);
}
private void Initialize(TimeSpan deltaBackoff, int maxAttempts)
{
this.deltaBackoff = deltaBackoff;
this.maximumAttempts = maxAttempts;
retry = new ExponentialRetry(this.deltaBackoff, this.maximumAttempts);
}
public IRetryPolicy CreateInstance()
{
EventExponentialRetry newInstance = new EventExponentialRetry(this.deltaBackoff, this.maximumAttempts);
newInstance.RaiseRetryEvent = this.RaiseRetryEvent;
return newInstance;
}
public bool ShouldRetry(int currentRetryCount, int statusCode, Exception lastException, out TimeSpan retryInterval, OperationContext operationContext)
{
bool shouldRetry = retry.ShouldRetry(currentRetryCount, statusCode, lastException, out retryInterval, operationContext);
if (shouldRetry)
{
OnRaiseRetryEvent(new RetryEventArgs(currentRetryCount, statusCode, lastException, retryInterval, operationContext));
}
return shouldRetry;
}
protected virtual void OnRaiseRetryEvent(RetryEventArgs e)
{
// Make a temporary copy of the event to avoid possibility of
// a race condition if the last subscriber unsubscribes
// immediately after the null check and before the event is raised.
EventHandler<RetryEventArgs> handler = RaiseRetryEvent;
// Event will be null if there are no subscribers.
if (handler != null)
{
// Use the () operator to raise the event.
handler(this, e);
}
}
}
See the CustomAzureStorageRetryPolicySample GitHub project for a complete sample.

Answer 2

You can make use of the same Enterprise Transient Fault Handling Framework with your Table Storage (and its associated retry notifications).
To do so you disable the default Table Storage transient fault handling and use the enterprise storage fault handling instead.
var retryPol = new RetryPolicy<StorageTransientErrorDetectionStrategy>(retryStrategy);
retryPol.Retrying += (obj, eventArgs) =>
{
var msg = String.Format("Retrying, CurrentRetryCount = {0} , Delay = {1}, Exception = {2}", eventArgs.CurrentRetryCount, eventArgs.Delay, eventArgs.LastException.Message);
System.Diagnostics.Debug.WriteLine(msg);
};
var options = new TableRequestOptions { RetryPolicy = new RetryPolicies.NoRetry() };
CloudTableClient tableClient = storageAccount.CreateCloudTableClient();
CloudTable table = tableClient.GetTableReference("people");
CustomerEntity customer1 = new CustomerEntity("Harp", "Walter");
TableOperation insertOperation = TableOperation.Insert(customer1);
retryPol.ExecuteAction<TableResult>(() => { return table.Execute(insertOperation, options); });