On 8/24/07, Jesper Krogh <[EMAIL PROTECTED]> wrote: > > > > I don't know if you can do Queing but from Jesper's description it > > didn't seem like order was important. Seemed to me that he just wanted > to > > stash something away and grok it later on. > > Thats pretty much the requirement. But it needs to be "fail-tolerant" and > survive a reboot of the computer without loosing messages. > > Jesper > --
If you require fault-tolerant (reliable message delivery) then you'll need to have a fail-over system for storage. Typically, this involves databases as Perrin posted earlier. If you do a Master-Master replication setup, you're guaranteed to have redundancy between the jobqueue and you won't have any loss of data. If you only have one server, you at least get the failure when trying to insert the job rather than it being lost in the ether. I'm not sure of the other messaging queues others have posted, perhaps they can address if the recommendations support this requirement. I can, however, say that TheSchwartz does work and is fault-tolerant (both in terms of hardware failures and recording if a message/job was unable to be processed properly). -J -- J. Shirley :: [EMAIL PROTECTED] :: Killing two stones with one bird... http://www.toeat.com
_______________________________________________ List: Catalyst@lists.rawmode.org Listinfo: http://lists.rawmode.org/mailman/listinfo/catalyst Searchable archive: http://www.mail-archive.com/catalyst@lists.rawmode.org/ Dev site: http://dev.catalyst.perl.org/