[infinispan-dev] FineGrainedAtomicMap preview

[Vladimir Blagojevic]

On 11-10-06 2:46 PM, Mircea Markus wrote:
>
> On 5 Oct 2011, at 22:58, Vladimir Blagojevic wrote:
> Why do we want to distinguish them? Is the FGAM significantly less 
> performant than AtomicMap? Otherwise why not just make all the maps 
> atomic.
You mean lets forget the legacy AtomicMap and make it fine grained by 
default?

> And there's an bug with the current non-fine grained atomic maps maps: 
> no WL is acquired when you write to such a map anymore. And it should 
> be, here's the UT. https://gist.github.com/1267310

Thanks!
> In order to understand the issue I needed to look into the entire 
> code. Here are some observations:
> - you shouldn't register a sync for applying the atomic map changes. 
> Why? There are several reasons for that, the main one being that with 
> this approach FGAM would't be able to participate in distributed 
> transactions.
> I don't see where the FGAM's entries that are added within a 
> transaction's scope are merged within the existing entries in the 
> FGAM. What do code does is keep all transaction-scoped entries in the 
> data container's FGAM instance, even if they are not committed.
> This test fails and it obviously shouldn't: 
> https://gist.github.com/1267303
Yes it fails. To be honest I did not think we need to support this. So 
for every delta change on AtomicMap I need to keep undo operation 
around? Manik and I completely overlooked this in our brainstorming.
> I think before moving forward with the review you'd need to change the 
> code to keep modifications in the transaction's scope and only apply 
> them during commit.
Right!
>
> Your test fails because on remote put you basically override the map 
> that exist on a node with the map that is obtained from the remote node.

You are right. Now I am sending only deltas never overriding one Map 
with another and the test passes!

Thanks Mircea!