near-cache interpretation

[ramki]

Hi,
We are trying to use near-cache and we are having many discussion
around the way it is supported in hazelcast, as the documentation was
a bit ambiguous.

1. Can we configure near-cache only for cluster members with data? or
for super client and/or native clients also? From my experience of
working on other caches, generally it is a concept of cluster members,
but i need to check to see if we should configure some of our nodes as
super client or native client. (as they don't need to be the owners of
data, but need fast access to elements which are accessed frequently)

2. To define <near-cahe> in spring, should this be on the instance
configs which will be the member or super client?

Appreciate the help.

Thanks,
Ram

[Mehmet Dogan]

1. Can we configure near-cache only for cluster members with data? or
for super client and/or native clients also? From my experience of
working on other caches, generally it is a concept of cluster members,
but i need to check to see if we should configure some of our nodes as
super client or native client. (as they don't need to be the owners of
data, but need fast access to elements which are accessed frequently)

Both normal (data) nodes and super-client nodes can have near-cache configured. Both are cluster members.

Native clients aren't. 

 

2. To define <near-cahe> in spring, should this be on the instance
configs which will be the member or super client?

 

You can configure near-cache on both member and super-client instance configs.

Take look at 'testMap' map configuration: 

http://code.google.com/p/hazelcast/source/browse/trunk/hazelcast-spring/src/test/resources/com/hazelcast/spring/fullcacheconfig-applicationContext-hazelcast.xml

@mmdogan