Spring Team
Glenn Renfro

Glenn Renfro

Spring Task/Data Flow Committer

Atlanta USA

As a Pivotal engineer, Glenn is a core committer for Spring Integration, Spring XD or wherever he is needed..
Blog Posts by Glenn Renfro

Spring Cloud Task 1.2.0.M2 is now available

We are pleased to announce that Spring Cloud Task 1.2.0.M2 is now available via Github and the Pivotal download repository. Many thanks to all of those who contributed to this release.

Spring Cloud Task 1.2.0.M2 offers the following features:

  • Upgrade to Spring Cloud Dalston RC1 and Spring Boot 1.5.2 - This release is compatible with the Spring Cloud Dalston RC1 release.

  • Fixed invalid data type bug in Oracle Script.

  • Fixed bug in MySQL migration scripts so that they will work on file systems that are case sensitive.

  • Updated Spring Cloud Task samples to use Spring Boot 1.5.2.

  • Updated documentation to discuss unit testing and notes on using Transaction Managers.

Read more...

Spring Cloud Task 1.1.2.RELEASE is now available

We are pleased to announce that Spring Cloud Task 1.1.2.RELEASE is now available via Github, Pivotal download repository, and Maven Central.

The main driver for this Spring Cloud Task release is to correct an error in the dependencies within the 1.1.1.RELEASE.

What do you think?

We look forward to your feedback on these new features in Github, StackOverflow, or to me directly via Twitter @michaelminella!

Spring Cloud Task Home | Source on GitHub | Reference Documentation

Read more...

Spring Cloud Task 1.1.1.RELEASE is now available

We are pleased to announce that Spring Cloud Task 1.1.1.RELEASE is now available via Github, Pivotal download repository, and Maven Central.

The main driver for this Spring Cloud Task release is to replace the MYISAM engine that is used to create TASK_SEQ table for the MySQL database with the InnoDB engine.

For more information please look at the following Github issue.

What do you think?

We look forward to your feedback on these new features in Github, StackOverflow, or to me directly via Twitter @michaelminella!

Read more...

Spring Cloud Task 1.1.0.RC1 is now available

We are pleased to announce that Spring Cloud Task 1.1.0.RC1 is now available via Github and the Pivotal download repository. Many thanks to all of those who contributed to this release.

Spring Cloud Task 1.1.0 offers the following features:

This release continues to progress on 1.1.0’s main theme of improving overall task functionality as well as migrating features that existed within Spring XD to Spring Cloud Task. The feature set for 1.1.0 includes:

  • Updated error message handling - 1.1.0 adds a new column to the TASK_EXECUTION table, ERROR_MESSAGE. This new column will be where any exception that is thrown causing the failure of a task to be recorded (this is changed from the EXIT_MESSAGE field in previous versions). This change allows for the EXIT_MESSAGE field to be better utilized for orchestration of tasks.
  • Improved options for customizing the launch of partitioned job workers - When using the DeployerPartitionHandler, 1.1.0 now provides abstractions around customizing the environment variables and command line arguments on a per partition basis.
  • Support for external execution id - When executing a task on a PaaS or other cloud infrastructure, there is typically a mapping that needs to occur between the task’s execution id and the id provided by the underlying system. This feature provides the ability to correlate those two for things like log management, audit, etc.
  • Support externalized task execution recording - This capability goes along with the support for an external execution id in that this allows the launching system to record a task execution’s existence prior to it’s start.
  • DB2 Support - Thanks to community for adding the DB2 support for Task Repository.
  • Partition Job Samples - Checkout the updated partitioned batch-job sample and the updated instructions on how to orchestrate them as short-lived Tasks in Local, Cloud Foundry, Mesos, YARN and Kubernetes.
Read more...

Spring Cloud Task 1.1.0.M2 is now available

We are pleased to announce that Spring Cloud Task 1.1.0.M2 is now available via Github and the Pivotal download repository. Many thanks to all of those who contributed to this release.

What’s new in Spring Cloud Task 1.1.0.M2

This release continues to progress on 1.1.0’s main theme of improving overall task functionality as well as migrating features that existed within Spring XD to Spring Cloud Task. The main updates for this release are the following:

  • Support for Generated Task Id - A common use case for a Spring Cloud Task application is to launch it within a PaaS, like Cloud Foundry or other cloud infrastructure. Because of this, there may be a delay between the launch request and the actual application beginning. Worse, the system responsible for launching the application may not actually launch the task for reasons outside of Spring Cloud Task’s control (lack of resources, network issues, etc). This feature allows the caller to register that a task has been requested to run. This way, no request of a task is lost between the request and the actual start of the Boot application.
  • Support for External Execution Id - When executing a task on a PaaS or other cloud infrastructure, there is typically a mapping that needs to occur between the task’s execution id and the id provided by the underlying system. This feature provides the ability to correlate those two for things like log management, audit, etc.
Read more...

Spring XD Benchmarks Part 1

Introduction

A common question when developing streaming applications is, “How many events per second can you process?”. The primary purpose of this blog post is to answer that question without falling into the classic benchmarking conundrum of benchmarking versus “benchmarketing”. The common approach with ‘native’ benchmarking applications provide by messaging middleware vendors is to focus on raw data transport speed, without serialization or deserialization of the message data and without any data processing. In Part 1 of the series, we will follow this approach.

Read more...