java - High Availability and Disaster Recovery Software AntiPatterns -
If it comes to high availability and disaster recovery, then you had to audit a Java application for the worst practices , Then possibly looking for hard-working IP addresses and sub-cell caching of bind handles. What else should be considered?
lack of verb / state logging.
A Java application should be able to resume when it crashed.
This means that a system that has already been able to record (next run in order not to repeat all)
This means that such a java program should always get the same status after the set of same tasks (doing something twice, the result of the same result, and already The work done should not be done again, but should be left only)
This record can take many forms (in a repository of file, database, matadata sort, ...), but The point is: the Java application can be as fast as a dip Quickly should be fine, it should know that he has already done.
Comments
Post a Comment