In the Facebook Memcache paper, we talk about dealing with transient failures by queuing up failed deletes to be replayed later: https://www.usenix.org/system/files/conference/nsdi13/nsdi13-final170_update.pdf
This works well if you are relying on deletes to keep your cached data consistent. If you don't use deletes, it becomes much harder to keep your data consistent and you have to think carefully through failure cases. Deletes are idempotent, so as long as we run each delete at least once, we know we'll end up in a consistent place. ~Ryan On Thu, Sep 11, 2014 at 10:25 AM, RamP <[email protected]> wrote: > Hi, > I am using distributed memcache cache. > > below is the scenario > i have 2 memcached nodes A and B > i have added an object obj1 with name '*John*' and it is stored in node A. > now obj1 need update, since it is not available in cache (as node A was > not accecible) we add obj1 againg with updated information now name is* > 'Peter'* and get saved to node B. > node A restored the network. > when i read obj1 from cache i am getting old value which is '*John'.* > > any one faced this issue?, apreciate any possible solution. > > Regards > > -- > > --- > 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.
