I need to implement a scheduler for running a periodic job. There is a table in a Cassandra database in which one column contains the creation time-stamps of rows and another column contains the status which is "open" by default. I need to create a job which should run periodically in every 10 minutes and query Cassandra to find out whether any rows with the "open" status are 5 or more days older. If so, the job should then change the status of all such rows from "open" to "closed". That's it.
The above seemingly simple problem is part of a large industry-strength OSGi application and I have been asked to pick the best 3pp option among:
1. There will be multiple instances of the application (and, therefore, multiple instances of that periodic job) running on multiple JVM's in the live environment interacting with the same Cassandra database with thousands of rows. [Do I need to handle concurrency from my Java code?] 2. I know Quartz does not guarantee persistence but do I really need it in my case?
3. Which of the 3pp solutions (Quartz, Sling, Zookeeper) can I use?