drizzle-discuss team mailing list archive
Mailing list archive
Re: Drizzle Replication Question
On Thu, Jul 16, 2009 at 9:17 PM, MARK CALLAGHAN<mdcallag@xxxxxxxxx> wrote:
> On Thu, Jul 16, 2009 at 11:54 AM, Jay Pipes<Jay.Pipes@xxxxxxx> wrote:
>> cc'ing the list since it may be interesting to others...
>> Padraig O'Sullivan wrote:
>>> Hey Jay,
>>> Here's a quick question on something I want to do that you might be
>>> able to help me with. For fun and because I think its a really
>>> interesting project, I'm writing a memcached applier which simply
>>> pushes update and delete events to a memcached server (for now).
>>> Basically, it enables someone to keep a pro-active cache easily.
> As someone who has reviewed all of the diffs for the replication
> changes made by Justin and Wei in the Google patches (limited fun was
> had by all who modified the code and reviewed the changes), my only
> response is .... wow.
> If making such a change in drizzle is easy and/or fun then drizzle is
> amazing and gives MySQL a chance of remaining relevant for large-scale
> datacenter deployments.
In case this is ambiguous:
'gives MySQL a chance of remaining relevant'
'gives MySQL a chance via Drizzle of remaining relevant'
Comparing MySQL with Oracle, Postgres and any other RDBMS is looking
backwards. There are many products in progress meant to out-scale
(HBase, Hypertable, ...) or out-avail (Cassandra, Voldemort, CouchDB)
the typical MySQL deployment. Enhancing replication will allow MySQL
to remain competitive in that market.
> Mark Callaghan