public async Task <LivyBatchResponse> ExecuteWorkflowAsync(LivyBatchRequest batch, TimeSpan waitTime)
{
LivyBatchResponse response = await Post <LivyBatchRequest, LivyBatchResponse>("batches", batch);
DateTime startTime = DateTime.UtcNow;
Exception lastException = null;
while (
(response == null) ||
(response.State != SparkJobState.Dead && response.State != SparkJobState.Error && response.State != SparkJobState.Success))
{
TimeSpan length = DateTime.UtcNow - startTime;
if (response != null)
{
Debug.WriteLine("session id: {0}, state: {1}, length: {2}", response.SessionId, response.State, length);
}
else
{
Debug.WriteLine("no response received, will try again");
}
if (length > waitTime)
{
throw new TimeoutException("the workflow has timed out", lastException);
}
await Task.Delay(TimeSpan.FromSeconds(10));
try
{
response = await GetBatchStateAsync(response.SessionId);
}
catch (Exception ex)
{
lastException = ex;
Debug.WriteLine("unexpected livy error: " + ex.ToString());
}
}
return(response);
}
/// <summary>
/// Executes a Livy job using a storage jar
/// </summary>
public async Task <LivyBatchResponse> ExecuteAsync(LivyBatchRequest batch)
{
LivyBatchResponse response = await Post <LivyBatchRequest, LivyBatchResponse>("batches", batch);
return(response);
}