About Me

My photo
"Enoughtheory.com" had its humble beginning in the year 2011 by ( Founder of Enoughtheory.com ) Mr Ravi Kant Soni , an Enterprise Java and Spring Framework Specialist, with a bachelor degree (B.E) in Information Science and Engineering from Reva Institute of Technology at Bangalore. He has been into the software development discipline for many years now. Ravi has worn many hats throughout his tenure, ranging from software development, designing multi-tenant applications, integration of new technology into an existing system, to his current love of writing a Spring Framework book. Currently, he is a lead engineer at HCL Technology. Ravi has focused on Web and Enterprise development using Spring Framework for most of his career and has been extensively involved in application design and implementation. He has developed applications for Core-Bank, HR and Payroll System, and e-Commerce systems using Spring Framework. Ravi Kant Soni is author of book "Learning Spring Application development" http://learningspringapplicationdevelopment.com

Wednesday 28 November 2012

J2EE Scheduler

Schedule-based execution in a J2EE Web Application

 
Schedule-based execution in a J2EE Web Application.
 
Scheduled tasks fall into two categories:
  • System level task, such as system backup. Usually, it is done by system administrators or operators at data center.
  • Application level, such as business report. Usually, this is the responsibilities of business units.
Usually, your application server is in your data center or at a hosting company. You want to control application tasks at your office on a client machine. Different departments want to manage their own tasks. For example, warehouses run inventory checking tasks while sales run sale analysis tasks. If anything goes wrong at the system level (for example database is down), the scheduler should send alert notifications to the administrator on duty. If you run scheduled business patrol tasks and find an abnormal condition at business level, the scheduler should send alarm notifications to the people in charge. All notifications should be handled by the scheduler automatically.

All the above issues require scheduler to support distributed computing environment with the following features:
  • The scheduler can run on any machine in the network at any time.
  • The scheduler can be run from data center and offices simultaneously.
  • The scheduled task should run once and only once at the scheduled time, regardless of how many of the schedulers are running.
  • Missed tasks can be detected and alerted. 

 

No comments:

Post a Comment