Bela, (cc'ing the jbosscache dev list) Sounds like a good idea and definitely easy enough to implement, but playing devils advocate, doesn't this completely undermine the meaning of "Synchronous Replication"? REPL_SYNC provides a guarantee that the method would not return until replication is complete to all cluster nodes. With this, that guarantee won't be there. Perhaps it should be a separate cache mode - REPL_SEMISYNC (*) ? Has this specifically come up as a request somewhere? (*) - terrible name!! On 27 May 2007, at 17:25, Bela Ban wrote: -- Manik Surtani Lead, JBoss Cache JBoss, a division of Red Hat Email: manik(a)jboss.org Telephone: +44 7786 702 706 MSN: manik(a)surtani.org Yahoo/AIM/Skype: maniksurtani