On Wed, Dec 9, 2009 at 5:43 PM, Jonathan Ellis <jbel...@gmail.com> wrote: >> In any case, at least for readRemote, why when the "suitableEndpoint" >> timeout another >> node is not tried ? > > Because > (1) clients need to be able to handle failures either way, so better > to bail early since > (2) each retry costs potentially RPC_TIMEOUT ms
Make sense, thanks for the answer. And wouldn't it be possible/reasonable to do something like a strong read, but with a modified quorumResponseHandler that return from get() as soon as it gets an answer and do the responseResolver/read repair in the background. Wouldn't it speed up some reads when a node other that the suitableEndpoint has less latency ? > > -Jonathan >
