On Wed, Apr 7, 2010 at 11:55 AM, dormando <[email protected]> wrote:
> > Usually it's a premature optimization or a flow problem if you need to do > that though. Real cases have been rare. > Yes, that's probably the case because it's not really using Memcached as a cache. In this case there's a number of processes on separate servers that share a cache entry to manage a non-critical and long-running task. The expires time is updated during this process. The expires time is used to remove the entry in the event the task fails. (It gets restarted by other means in that event.) If the task is completed then the key is removed. The question came up if there was a way for another process that is only aware of the key (has no knowledge of the structure of the value) to "ping" memcached to see if the the key exists w/o having to fetch the data, and to determine when to "ping" again based on the expires time. Yes, it's abusing the purpose of the cache a bit -- not the first time that's happened. ;) Thanks, -- Bill Moseley [email protected] -- To unsubscribe, reply using "remove me" as the subject.
