tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Chong Yu Meng <>
Subject Re: JDBC & ORACLE implementation !
Date Thu, 20 Feb 2003 09:11:35 GMT
Hi Swapneel !

I need you to clarify something : are you processing the files first, 
then putting them into the database, or are you putting all the raw data 
inside the database and doing SQL operations to extract the data you 
want ? I am assuming that you want to do the second option, am I right ? 
I need to make a few other things clear :

1. If you are committing 1MB of data into the database per operation, 
that would take a looong time, and I'm not sure if that is the most 
efficient, or even the easiest way to do what you are trying to do. Why 
I say that is, Oracle is fine when you break up the data and commit, 
perhaps line by line, to the database. Not very fast or efficient, but 
still manageable. My experience of Oracle, though limited, has not 
exactly given me much confidence in its ability to handle huge data in 
individual fields. Stuff the fields with too much data, and try to work 
with them, and you will realize the necessity to tune Oracle. I'm not an 
expert on Oracle, nor do I care to be, so I won't try this.

2. String manipulation with Java 1.4 is possible and quite easy, now 
that it supports regular expressions. If, like me, you'd like to do 
things the "hard way", you can use the string handling functions in Java 
to do the processing for you. It's actually a lot easier than it sounds. 
I refer to O'Reilly's Java Cookbook for examples on this.

3. The least risky way (and some would say, the lazy way) is to just 
improve on the existing code -- use the FILE SYSTEM solution that 
already exists and extend the functionality, fix bugs, or improve the 
GUI. It's not because you may be "re-inventing the wheel", as some may 
say, if you choose to implement it in a database. I am assuming that you 
don't have a lot of time, and there is no in-house expertise you can 
refer to (i.e. programmer friends where you live, that you can ask), 
and, as you say, you are coming into this for the first time. As a rough 
indicator (and I invite others to correct me), it takes :

- 1 month to understand Oracle
- 2 weeks to get JDBC working the way you want
- 2 weeks or more to work out the business logic
- 1 week to figure out how to draw graphs
- 2 weeks to fix bugs
- 1 week to understand the deployment environment
- 1-2 weeks to fix problems related to the deployment environment
- Total : 14 weeks

That's more than 3 months, and assumes that you devote yourself 100% to 
the project. A filesystem solution however takes :

- 1 week to learn about file i/o
- 1 week to learn how to do string manipulation
- 1 week to learn graphs
- 2 weeks to fix bugs
- 2-3 weeks to deploy
- Total : 8 weeks

You might want to just go with a simple file system solution for now, 
and slowly work out how to do the rest later, when you are not hurried. 
It really is better.

pascal chong

Swapneel Dange wrote:

> hey CHONG !
> yepp, i think i am hurrying too fast, and i have the FILE SYSTEM with 
> me but the thing is that i need a sytem is which i can atlesat shoud 
> be able to do something like this --
> 1) strip of the files of the un-necessary content.
> 2) then use the data for searching patterns.
> 3) apply the patterns to the unsual causes.
> 4) then use the data to relate to GRAPHS
> right now i am using a file system in which i cant perform the string 
> matching and other operatoins. moreover all thsi has to be implemented 
> on teh WINDOWS NT or XP. so i am sure as to how u implement somethign 
> like shell scripting in windows. anyways thats more or less knowledge 
> i ahev about al lthis.
> will be great to hear from u about all this ! thanx !
> Swapneel Dange
> 505-642-4126

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message