Downloads of v 0.2.0.16
Service Bus Durable Task Framework
This package works with Windows Azure Service Bus and provides a C# based durable Task framework for writing long running applications. It adds Microsoft.ServiceBus.DurableTask.dll along with related libraries to your project. Additional info for Durable Tasks is at http://aka.ms/l7283l,
Please note that this package requires .Net Framework 4.5.
To install Service Bus Durable Task Framework, run the following command in the Package Manager Console
PM> Install-Package ServiceBus.DurableTask
- Added OrchestrationStateQuery user model with in-memory implementation
- Sorted results returned by TaskHubClient.GetOrchestrationHistory() by event timestamps
- Added pagination for OrchestrationStateQuery results
- Added async methods to TaskHubClient
- OrchestrationContext.ContinueAsNew now has an overload that takes the desired version of the new generation
- OrchestrationContext.CreateSuborchestrationXXX methods have overloads that take an explicit Instance Id to use as the instance id for the suborchestration
- Fixed null ref in tracing when terminating an orchestration
- Added more resilience against Service Bus clock skew
- Fail the orchestation on detection of non-deterministic workflow
- Remove explicit retries for service bus operations (sb client does that)
- Log critical error if delivery count if it gets near the maxdeliverycount
- Terminate orchestration with appropriate message if session size grows above 230k.
- Added backoff on messagingexceptions even while processing work items
- Increase Max delivery count on taskhub queues, make it configurable.
- Log message stats if tracking dispatcher write fails
- Abandon messages properly in the tracking dispatcher (similar to what we do in the orchestration dispatcher)
- Added TaskHubWokerSettings option that gives users access to concurrency of TaskOrchestration, TaskActivity and Tracking dispatchers
(concurrent items + backoff etc)
- Orchestration state compression can be specified via the TaskHubWorkerSettings object. Turned off by default. This helps write more complex orchestrations
without running into the session state size limit.
- Fixed bug in Tracking dispatcher which causes failure when dumping a history event larger than 32K to a column in the Azure table store.
- Fixes a critical regression in 0.2.0.13 that caused sessions in service bus to never get deleted even when orchestrations were terminated. Also added
tests to catch this regression in the future.
- Added TaskHubWorker.CreateHub(bool createInstanceStore) overload that would let users control whether they want to drop and recreate the instance store
along with the task hub queues.
- Adds message level compression and associated options to control compression and threshold in TaskHubWorkerSettings and TaskHubClientSettings.
- Removes batching for Send operations within the TaskOrchestrationDispatcher's transaction block. Reason was that message batches are bundled up into a single
batch message which has to follow the 256KB message size limit thus limiting the response size for task activities or the complexity of the orchestrations.
- Fix bug with RetryOptions when TimeSpan.Zero is passed in as firstRetryInterval
- Fix bug with tracking dispatcher failing to truncate larger history entities when writing to the Azure table based instance store
- Added a PurgeOrchestrationInstanceHistoryAsync method on TaskHubClient which enables users to purge older orchestration instance histories from the instance
store. This relieves pressure on instance queries.
|Service Bus Durable Task Framework 0.2.0.16 (this version)||21||Friday, July 11 2014|
|Service Bus Durable Task Framework 0.2.0.5||1051||Friday, January 24 2014|
|Service Bus Durable Task Framework 0.2.0.2||128||Thursday, October 17 2013|
|Service Bus Durable Task Framework 0.2.0.1||127||Saturday, September 21 2013|
|Service Bus Durable Task Framework 0.1.0||132||Wednesday, June 26 2013|