That might be done in the client. You can turn on replicating and
fail-over, during fail-over you need
read-repair to restore data on the resumed instance. However  to make
these work correctly, careful programming is needed.

Best regrads,
Chen,Congwu



On Thu, Apr 5, 2012 at 10:04 PM, chmin wu <[email protected]> wrote:
>      Hi, I have two memcached server running in different machine for HA
> purpose- they are used to backup for each other. The data stored in
> memcached is very important for my application; memcached servers must have
> the same data for backup.
>
>     So the question is,  if a memcached server crashes and resume, how to
> completely replicate all the data from the  the healthy one to the rusumed
> one?
>
>     Many thanks in advance!

Reply via email to