Re: [infinispan-dev] JGroups 2.7 and 2.8 work with Infinispan now
On 1 Jun 2009, at 10:22, Mircea Markus wrote:
Bela Ban wrote:
>
>
> Mircea Markus wrote:
>> Vladimir Blagojevic wrote:
>>> Hi,
>>>
>>> Give it a try as well! I suggest we move to 2.8 sooner than
>>> later. Currently, with tcp and 2.7 I get 0 failures and with 2.8
>>> I get 9 faliures.
>> That's something!
>> The way I see it we are kind of trapped: on one hand we cannot fix
>> NBST without a JGroups fix,
>
> I thought NBST doesn't use JGroups anymore ? (I mean the block()
> and unblock() and flush functionality)
that was a hack[1] in order to fix a race condition taking place
when a new member joins.
It didn't work, though.
Also, this is not specific to NBST as it affects normal traffic as
well. From a direct email between Mircea, Vladimir and myself, the
following sequence could result in a race even without NBST:
Could we create a test where:
1. Node A and B are in the cluster.
2. Node A does put(x, y) the same time Node C joins
3. Test that Node C sees the put.
4. Repeat the test to simulate the race condition.
I believe the problem is that C discards messages before its view is
installed. A & B could receive this view and send messages to C,
which C will not see. Initially we thought FLUSH will work around
this problem, and while it reduces the window of occurence, the
problem still exists since there is a race between A & B receiving
unblock() and then broadcasting queued messages, and C receiving these
queued messages and receiving unblock().
One solution which Mircea proposed is for FLUSH to queue messages
received during a blocked state, so that this race condition on C is
removed. However it would be better to do this without FLUSH, for C
to queue messages received between joining and its view being installed.
Cheers
Manik
>
>> and JGroups fix can only be performed on 2.8, and we cannot go
>> live with it.
>> When is 2.8 scheduled for going GA?
>
> I think by the end of the summer. I'll release a first beta soon.
> 'Beta' sounds scary, but as a matter of fact 2.8 is very stable
> already.
>
>
[1] https://jira.jboss.org/jira/browse/ISPN-83
_______________________________________________
infinispan-dev mailing list
infinispan-dev(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Manik Surtani
manik(a)jboss.org
Lead, Infinispan
Lead, JBoss Cache
http://www.infinispan.org
http://www.jbosscache.org