[KLUG Members] Postgres Cluster

Adam Tauno Williams adam at morrison-ind.com
Tue Jul 19 13:41:50 EDT 2005


> > > adequate locking.  That would mean that each "peer" would have
> > > to get an acknowledgment from each of the other peers that that
> > > it has received a "record 53 of table mytable" is locked. 
> > Yes, and implement START/STOP operation notificaiton with
> > timeouts, post-fault resyncronization, and multi-host rollback. 
> > Yep.  And you wonder why almost no one support this? :)  Of
> > course, almost no one needs it.
> > You run the databases on high-end machines, most likely with
> > fiber-channel interconnects. $$$,$$$.99 at a minumum.
> MySQL Cluster does exactly this;  your application connects to a
> single host machine, but only when all machines in the cluster
> have commited the entire transaction does the host return the
> status to your application.  Very tricky to get right, but they've
> done it.
> They do recommend special NICs designed for low latency, for
> obvious reasons.  Not that expensive, maybe $1K a card, SCI is the
> name I've heard.  But the DB machines can be low five figures
> (each) or even high four figures... Opterons are cheap nowadays.
> And Emic offers a very similar functionality, based on MySQL,
> without requiring RAM residency.  <http://www.emicnetworks.com/>

m/cluster -
"Emic™ m/cluster is the leading application clustering solution for
MySQL.  m/cluster supports clustering for all the most popular MySQL
storage engine types: MyISAM, InnoDB and memory (heap).
m/cluster delivers the robustness that you can rely on to build
mission-critical applications using the world’s leading open source
database, MySQL. Your requirements for high availability and performance
scalability are met by adding m/cluster to your application. 
m/cluster creates a single virtual database view of multiple MySQL
databases and is designed to integrate seamlessly with your MySQL-based
application.
The architecture of m/cluster is designed to use commodity hardware and
does not require any special hardware devices like storage area networks
or load balancers."
- sounds like a true "cluster" to me.   It even seems to guarantee
atomic operations.

The - "All cluster nodes share the same IP address, which makes the
whole cluster appear as a single IP network element"  - is a neat
approach. 

> > I know of an Informix cluster instance of two geographically
> > distant servers interconnected by an OC-3.
> Right, if your application is important enough that you need data
> integrity and failover during a tornado, earthquake, etc., then a
> geographically separated cluster is an expensive but viable
> solution.

Yep.



More information about the Members mailing list