All the logic you're describing is up to the client implementation; the servers just respond to requests they receive.
Some clients might failover a get to another memcached server; others might return an error. Clients like mcrouter have configurable options on how to behave. Probably many other clients are configurable as well. ~Ryan (mobile) > On Dec 4, 2014, at 8:52 AM, Gerard Ottaviano <[email protected]> wrote: > > Hi, new to memcached. When a memcached server (#1) goes down and a > subsequent fetch from cache to that server (#1) results in a miss and that > results in a fetch to the backend, does that now create a new cache entry on > one of the other memcached servers (#2) that is still available? If so, when > the down server (#1) becomes available again, and the next fetch of that same > entry from the cache is requested, how does it determine from which server to > get it, now that it exists on 2 servers? Do we need to save the hashes which > contain the server on which that cached entry exists, so we get the most > recent one? Thanks. > -- > > --- > You received this message because you are subscribed to the Google Groups > "memcached" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. -- --- You received this message because you are subscribed to the Google Groups "memcached" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
