Hi,

We recently noted an issue where a ConcurrentHashMap collection used within
MB for storing undelivered messages seemed to return corrupt data when
calling the "remove" method (null objects).

A fix [1] was done (pending merge to wso2-dev) as follows :
An extension class was created and used, adding ReadWrite locks to
ConcurrentHashMap operations.
"get" method was used to retrieve the data, after which the "remove" is
called.


Have you come across any similar problems when using ConcurrentHashMaps ?

Could there be other explanations to receive corrupt data in a concurrent
application when using thread-safe collections ?

refer link for more details on the issue.

[1] : https://github.com/wso2-dev/andes/pull/24

Cheers,
Hasitha Amal De Silva
 Software Engineer
Mobile : 0718133463
WSO2 Inc.: http://wso2.com ( lean.enterprise.middleware. )
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev

Reply via email to