I had a look, still have a few concerns:
* Why do we still have DistributedTaskContext?
* Factory should probably not be genericised. E.g.,
interface DistributedCallableFactory {
DistributedCallable<K, V, T> createDistributedCallable(Cache c,
EmbeddedCacheManager ecm);
}
but then again that will depend on what you think re: my proposal to make this look like
an ExecutorService (see separate email)
The M/R stuff looks good. :)
Cheers
Manik
On 13 Jan 2011, at 16:40, Vladimir Blagojevic wrote:
Hey Manik and all,
https://github.com/vblagoje/infinispan/tree/2ce49c49ea2cb8ba8d0e7d3e4829b...
Let me know if you have comments. The plan is to go forward with this
proposal for ALPHA2 tomorrow.
Regards,
Vladimir
On 11-01-12 2:42 PM, Vladimir Blagojevic wrote:
>> Also, wrt. names and methods on the API on the Distributed Exec part
>> of things, wdyt about aligning it with JDK executors (see separate
>> email from me)?
>>
> What exactly are you referring to? API like this one:
>
> DistributedForkJoinPool p = cacheManager.newDistributedForkJoinPool();
>
> result = p.invoke(new DistributedTask() { ... });
>
>
>
> Do we need an executor service? We invoke tasks directly with execute
> and executeAsync while MapReduceTask is invoked with:
>
> R result = new
MapReduceTask(cache).on(K...).mappedWith(mapper).reducedWith(reducer).collate(collator);
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
twitter.com/maniksurtani
Lead, Infinispan
http://www.infinispan.org