----- Original Message -----
> From: "Charles Brown" <CBrown@stripped>
> Anyone out there with experience in Mysql Clustering. My management
> requests that i migrate from replication to clustering. Why? Because
> everyone is doing clustering and he would like to stay competitive.
Your management has no clue what it's talking about. Shoot it in the knee and carry on.
Free business advice: if you believe randomly changing the technology underlying your
business is what it takes to stay competitive, hire someone who actually knows how to run
> The question is what are the pitfalls -- if any? Our replication
> objective is to address availability.
Pitfalls, like, you're working with a totally different technology that will handle both
data storage and retrieval entirely differently from what you now have?
For one, the SQL interface to MySQL cluster is merely a compatibility layer, if you want
to really reap the benefits from it you need to switch to NDBAPI.
You now have (probably) two servers; don't even *think* of starting with Cluster with less
than three servers, and make sure they have as much memory as possible, ideally more than
the size of your dataset.
You will also face different restrictions on data types, indices and whatnot. NDB is an
entirely different engine from MyISAM and InnoDB - on some fronts it's actually closer to
NoSQL than to SQL.
Don't get me wrong, it *could* be a good choice, but don't just switch because some dumb
manager is chanting buzzwords.
First of all, have your manager formally identify the problems there are with the current
setup, under the flag of if it ain't broken, don't fix it. If it merely comes down to "we
want more redundancy", have a look at SchoonerSQL, Clustrix or ScaleBase - those are
third-party [mostly|fully]-compatible MySQL forks with emphasis on multimaster
Bier met grenadyn
Is als mosterd by den wyn
Sy die't drinkt, is eene kwezel
Hy die't drinkt, is ras een ezel