CacheKey implementation doesn't support distributed caching because it includes
an internal object hashcode
-----------------------------------------------------------------------------------------------------------
Key: IBATIS-555
URL: https://issues.apache.org/jira/browse/IBATIS-555
Project: iBatis for Java
Issue Type: Bug
Components: SQL Maps
Affects Versions: 2.3.3, 2.3.2
Environment: Windows XP
Reporter: Sijo Mathew
The current implementation of CacheKey uses the native object hash code of
internal iBatis objects at some places to update the Hash Key. Since those
internal object hashcodes will be different between JVMs, even when executing
the same query/parameters, integration with a distributed cache is of very
limited value. To see this, imagine a distributed system with 8 client JVMs
and a central distributed cache (eg. oscache). While each client may execute
the same query+params, they will each generate a different CacheKey when
searching the cache. Therefore, a cache entry entered by JVM 1 will not be
visible to JVM 2 even if they have the same query+params and will return the
same result set. In general, different JVMs in a distributed system will be
unable to share cache entries.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
