Jennifer --

My company is putting a 3-node Cluster into production within the next 3 weeks. Right now we 're porting our application code from 3.22 to 4.1. Our servers won't be heavily loaded--we're switching to NDB for survivability, not scalability. We didn't purchase the JumpStart option.

It's surprising to me that more organizations haven't put MySQL Cluster into production. My guess is that 1) most are taking a wait-and-see approach; 2) nearly everyone has a working, dependable solution--with replication or whatever--and if it ain't broke, don't fix it; 3) some particular limitation of the Cluster product (e.g. only 128 columns allowed per table, no indexes on text fields) prevents them; and that the #1 reason is the expense of fulfilling the RAM requirement. Probably people are waiting for version 5 and wishing for an on-disk option.

As for our company, we feel the Cluster as is suits our needs, and after 9 months of testing--and 6 months of working with EMIC before that--are pretty eager to put it into production.

I'll let this list know how successful we are ;)

--- Jennifer Goodie <jgoo...@att.net> wrote: