[opencms-dev] Database connection pooling

[Evans, Michael]

Vadim,

1.  Not all app servers provide connection pooling.  Also the where class
names vary between implementations OpenCMS would then start to become
product specific.

2.  The structure of the SQL commands varies between different RDBMSes...

3.  Probably required for the whole content management, templates etc stuff
to work.  WAP - create a template that produce WML instead of HTML.

4.  Scaleability of end users visiting sites is excellent (with options to
static export content for greater scaleability), scaleability of interactive
users modifying content is moderate.


Regards,

Mike

-----Original Message-----
From: Vadim Lotarev [mailto:vlotarev at evelopers.com]
Sent: 20 February 2003 13:54
To: opencms-dev at opencms.com
Subject: [opencms-dev] Database connection pooling


Hello everybody,

I just review OpenCMS to investigate whether it is possible to use it in my
projects. There are some questions:

1. It seems that OpenCMS uses it's own database connection pool. Why don't
rely to datasource connection pools provided by any of production
application servers?

2. As you use SQL queries only why do you need specific settings for
different databases (except specific JDBC driver class and connection
string, of course)? BTW, why don't you use Datasource interfaces instead of
DriverManager-related methods?

3. Why using of own classes to convert XML to HTML is better than using of
XSLT technique? In what way do you support WAP (it is proclaimed in the
features but I am not able to find it in sources). Is it necessary to extend
sources to support any other convertations?

4. What about OpenCMS performance? I am not able to find any information
concerning this.

Thanks in advance,
Vadim Lotarev