[infinispan-dev] L1 Data Container

Mircea Markus mmarkus at redhat.com
Wed Jul 10 08:08:47 EDT 2013


On 9 Jul 2013, at 19:06, ben.cotton <ben.cotton at jpmorgan.com> wrote:

> /> This would only affect the Cache internals (use two DataContainers instead
> of one). Users still won't have direct access to the L1 cache.
> /
> 
> This comment prompts me to ask the Infinispan DEV team if they believe the
> following potential user API additions have merit:
> 
> Given that a user's  DIST_SYNC Cache can register a Listener to process a 
> 
> *CacheNotifier.notifyCacheEntryVisitedEvent()*
> 
> callback,
> 
> 1.  provide an API enhancement so the user's  Listener can identify the
> specific node at which the Visit occurred.

Possible to do it now: event.getCache().getAdvancedCache().getRpcManager().getAddress() 

> 
> 2. provide an API enhancement so the user's Listener can identify if the
> Visit occurred at the L1 near cache.

You can determine if the key for which the call was performed is local to this node:
event.getCache().getAdvancedCache().getDistributionManager().isLocal(key).

> 
> I'm just thinking that it would be nice if a user could see a /just few
> more/ exact details of how his/her
> 
> *V = cache.get(K);*
> 
> invokes distribute as "hits" on their DIST_SYNC Cache's  grid.

If you're interested for this from a statistics perspective, you might want to take a look at ISPN-2861[1], (documentation to follow) which adds a lot of interesting statistics to ISPN.


[1] https://github.com/infinispan/infinispan/pull/1938 

> 
> Thanks,
> Ben
> 
> 
> 
> 
> 
> 
> --
> View this message in context: http://infinispan-developer-list.980875.n3.nabble.com/infinispan-dev-L1-Data-Container-tp4027447p4027632.html
> Sent from the Infinispan Developer List mailing list archive at Nabble.com.
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev

Cheers,
-- 
Mircea Markus
Infinispan lead (www.infinispan.org)







More information about the infinispan-dev mailing list