[olug] Real opportunity for getting Open Source into the Enterprise

[Jeff Hinrichs]

In my experience most of these questions can't be answered until we have
som e idea of the size of the current DB.  What the standard growth rate
is, so we can forecast data storage requirements for a 5 year time frame.

Also, we would need a basic functionality list to know what to ask next
AND it would be nice to know where/how the data is currently stored.  If
it's win16 then dbase/btrieve are the most likely answers.  Which would
tend to make me think that Oracle would be overkill.  Once we know
functionality and data requirements we can start debating
mysql/postgres/et al.  For a volunteer project OSS though, I would think
that we would use only OSS components.

-jeff hinrichs

Nick Walter said:
> I write this sort of thing for a living nowadays (for telecoms to
> track/bill subscribers) and java servlets + jsp pages really are the way
> to go for front-end access in my experience.  The only downside to
> servlets is that the memory/cpu overhead of tomcat is not database
> friendly.  For decent performance you need 2 machines dead minimum, one
> for db and one for apache/tomcat.  Plus desktop machines with web
> browsers for access of course :)
>
> I've only ever worked with Oracle as a database back-end, and that's way
> too high cost for this scenario.  Anyone got experience with any of the
> free databases, are they robust enough for this type of application?
>
> Nick Walter
>
>
> On Mon, 2002-11-25 at 23:50, Daniel Pfile wrote:
>> On Monday, November 25, 2002, at 11:02 PM, Andrew Holm-Hansen wrote:
>>
>> > I don't know PHP but I know enough Java to get a servlet container
>> to  do the job.  I've gotten a sort of similar app working using a
>> Postgres backend and Tomcat as the servlet container.
>>
>> The reason I brought up php is most people know it, and it's low
>> system  requirements. Java servlets would be fine as well, I've never
>> written a  large project in it, but I've got all the books and I'd
>> love to give it  a shot. I actually debugged some servlet code (not
>> mine) tonight even,  so maybe I'm underestimating my skills...
>>
>> Also, a friend of mine who's a big delphi hacker (I know.. I know..)
>> Is  looking to get into java, since he realizes delphi isn't a way to
>> maintain marketability. I bet he'd love an excuse to hack out some
>> java  code for the experience of it. I'm not much a 'traditional'
>> programer  (more of a sysadmin) but I love programming, and would pick
>> up the  project for similar reasons.
>>
>> > I've also done some data conversion, though depending on the schemas
>>  and the amount of data to be converted I may decide to punt on this
>>  particular project.
>> > I'm willing, and I may even be able!
>>
>> Agreed, the scale of the project and the state of the old data is a
>> big  sticking point for me. I did a conversion project a few years
>> ago, and  the biggest problem with flaws in the way the data (old dbf
>> files) was  set up. It worked great on the old system, and great on
>> ours, till we  tried to bring up record 12,384 for example and found
>> out it was  corrupt, and there were lots more that way...
>>
>> Sourceforge could give us a nice place to set up cvs/docs/etc, and I'm
>>  sure all of us could set up a tomcat/postgres setup on our respective
>>  development machines to test on.
>>
>> -- Daniel
>>
>> _______________________________________________
>> OLUG mailing list
>> OLUG at olug.org
>> http://lists.olug.org/mailman/listinfo/olug
>>
>
> _______________________________________________
> OLUG mailing list
> OLUG at olug.org
> http://lists.olug.org/mailman/listinfo/olug