What is the better way to use a Jackson serialized in multithreaded system:
- To create a new serializer object per each request, or
- To create a serializer factory as a bean and make all requests to use it centrally
Does it have a synchronized blocks in serialization/desalination process, that can impact on runtime performance?
See http://jira.codehaus.org/browse/JACKSON-570?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel. It depends on what version you are using.