On the mirroring data are we talking about either doing DNS with a DNS->HTTP
gateway or dumping the data and allowing others to run the import command on
it? Or having the master generate whatever text files will be used behind
the query engine and distributing them to the mirrors, saving the DB dump
for those who want to play with all the extra junk?
Time for me to start putting together some toys of my own....
August
-----Original Message-----
From: Alan Cox <alan nospam at redhat.com>
To: cdindex nospam at freeamp.org <cdindex nospam at freeamp.org>
Date: Wednesday, March 10, 1999 1:27 PM
Subject: Re: My thoughts on this issue
> So where does this leave us? Who knows... I suspect we are all going for
> clean, simple, fast, flexible...
Yep - which really rules a relational database out for the export back end.
You are right that its definitely the tool to generate the master databases
and to handle the exporting. But the query engine should be running faster.
Thats why we have LDAP not SQL for big use information bases for example.
Oh I've got a new toy btw. This assumes you mirror the data some other
way but its now doing over 4000 queries/second via a small udp messaging
agent
Alan