maria-developers team mailing list archive

["Jakob Lorberblatt" <jakob@xxxxxxxxxxxxxxxxxxxxxxx>]

Hello,

I have not been very active in my open source (or online!) life lately as
I have been busy with my own wedding and starting work programming for a
new company. I apologize for my freebsd buildbot not having been up, a
power outage and a software raid issue; followed by the afore mentioned
wedding have kept me from bringing it back online or otherwise
pre-occupied so I hadn't stopped to think about the buildbot not having
started on boot. My apologies..

All of this aside I am about to launch a new software product; well more
of a production its for a single client and is not intended to come in a
box at this moment (or ever, should be freely available once I can get
organized enough again to get a project page together). Its for
inventory/calendar and truck management for an HVAC company; I would like
to base it on MariaDB explicitly but with enterprise class stability and
support forward looking. I have a good idea already of how to use the
database itself however I would like a fail safe (though small scale,
think small business) which is low in hardware requirements in general (to
make redundancy more affordable, mostly).
The big topic is, storage engine, I am actually still seeing performance
gain in many simpler scenarios from the classic MyISAM moniker from recent
applications I have launched; also the straight up file copy type backup
being an option is appealing for simplicity in scenarios where the user(s)
are less savvy but definitely dependent on their apps. Is there a
reliability loss with PBXT compared to InnoDB/XtraDB?
    I would like to do this right, in addition I do plan to make this
community information with the site I am launching. If anyone could
outline why one engine is better over another for this simple
application that would be great. I would like to build something very
progressive and supportable for this client who is personally
important to me. Also Language choice I suppose would be a factor,
python and java have suited me well and seem to due extremely well
interfacing with MariaDB performance wise with widely available stable
MySQL/MariaDB API's.
    Lastly, I will make it brief, Postgresql interoperability? My current
employer uses this database extensively; I would like to replicate
(even if only by snapshot) there data structures into this environment
for development, I believe in the long run a performance gain may be
attainable this way, is there any outlook of something like FederatedX
(or other method) allowing interoperability; replication would be
ideal (especially since postgresql does have some short comings as far
as replication methods), slony is currently employed. Our software is
from a significant consulting firm who has informed me that it should
work on basically "any" dbms. Replacement isn't an option at this
point in the deployment but tools to allow the two backends to work
together would open some doors for me, hence bringing up federation.
This is an old topic I know but to have tables that I are dissparate
but centrally acessiible by SQL via MariaDB would be preferrable and
very neat.
  Thank you for your time, and I would appreciate any feedback that I
receive.

Jakob Lorberblatt