[ale] MySQL Cluster

Michael Trausch mike at trausch.us
Sat Feb 18 15:05:33 EST 2012


You need some way to take out an atomic lock... or, use a message queue.

E.g., one system processes the new things and inserts their data into a
message queue, and you can then have N workers (where "N" is "as many as
needed on average) to pop things off the queue and send mails based on the
content of the message.

--
Sent from my Ice Cream Sandwich!
On Feb 18, 2012 3:00 PM, "Chris Fowler" <cfowler at outpostsentinel.com> wrote:

> On Sat, 2012-02-18 at 09:22 -0500, Mike Harrison wrote:
> > In reality, it looks like we'll just have to configure a cluster
> > and do some experiments. I know we'll have to code around some
> > potential
> > issues with unique keys.. and figure out other management issues.
> >
> > Like any complex system, the answers get more specific the more
> > complex the scenerio is.
> >
> >
>
> Well then experiment with me.  I've been wanting to do this, but I need
> to rework the applications so that I don't have them doing all the same
> thing.
>
> I have one application that looks for new rows in table A and then sends
> out email.  If I have a cluster of these servers. each running that app,
> how do I keep them each from sending out the same email?
>
> My issues are more application issues than getting clustering to work.
> I'm sure I can do master-master replication.  I need to get the programs
> to behave.
>
> Chris
>
>
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.ale.org/pipermail/ale/attachments/20120218/dd52ceac/attachment.html 


More information about the Ale mailing list