Döhr, Markus ICC-H schrieb:
does anybody know how MaxDB performes when one creates lots
(> 100) of rather small (< 200 MB each) databases on a single server?
I think you will sooner or later reach operating system imposed limits
(depending on the OS you're using) because each DB will (also depending on
the configuration) create 20 - 50 threads for each, additionally you will
need x_servers/vservers for the connections, so the scheduler will need to
deal with 2000 - 5000 threads and some hundred processes.
Hm, seems to be a bad idea.
Why do you want to create so many small instances? Just curious...
We are running a shop engine (build in perl) that uses MaxDB as its
database. At one of our customers the system uses a single
database-instance for about 120 shops utilizing a model like SAP does
(client-dependent and client-independent tables).
Now we migrate the system to java (using spring and hibernate) and are
focused with the problem that compound primary keys which are a
consequence of using a client-field in nearly every table are not realy
a good idea in such an environment. Hence the idea to split the clients
to different database instances...
Cheers,
Dirk
-------------------------------------
Lachowski Datensysteme GmbH
Strassburger Platz 25
D-30853 Langenhagen
Fon: +49 511 725353-71
Fax: +49 511 725353-89
--
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
