Desing Question related with Lucene

[ilkay polat <po...@gmail.com>]

Hello;

  I have a desing question while developing my project. If you have time,
  lease read my problem and if you have a solution please make me informed.

  Project : Our system produce a txt file for every one hour(13 pm 14 pm
e.g.
). (These files contain logs from network e.g. TCP logs). I use FreeBSD and
  cron. For every hour, after five minutes later(13:05 pm , 14:05 pm e.g.),
  there is a process which indexes this txt file with lucene indexing.  Then
I
  have an web app which search some textual search with this produced
indexed
  file(lucene produce obviously).

  Problem: Our customers wants to know which client is using internet most,
  which site is used most and like this things which are done with sql like
  that as you know

  select site, count(site) from log_table
  group by site

  Mhy solution is: A second process which insert logs to table (temporary
  table ) and aftr inserting do some quesries on this temporarty table and
get
 results to main statistics tablse which has tables as which site is most
  visited table . This temporary table updates related statistics table .


  I have need a recommendation about the problem? Is there any solution on
  lucene(get most ranked kind query searching is exist or not and if yes is
it
 good for performance)
  If there is no a solution in lucene what will you use for this situation?
: Thanks for your help.


: ilkay POLAT
: Research &Development Software Engineer TURKEY

Re: Desing Question related with Lucene

[ilkay polat <po...@gmail.com>]

Is it better to analyze logs with lucene ? Or other solutions are better for
performance....

On Thu, May 20, 2010 at 9:51 AM, ilkay polat <po...@gmail.com>wrote:

> Hello;
>
>   I have a desing question while developing my project. If you have time,
>   lease read my problem and if you have a solution please make me informed.
>
>   Project : Our system produce a txt file for every one hour(13 pm 14 pm
> e.g.
> ). (These files contain logs from network e.g. TCP logs). I use FreeBSD and
>   cron. For every hour, after five minutes later(13:05 pm , 14:05 pm e.g.),
>   there is a process which indexes this txt file with lucene indexing.
>  Then I
>   have an web app which search some textual search with this produced
> indexed
>   file(lucene produce obviously).
>
>   Problem: Our customers wants to know which client is using internet most,
>   which site is used most and like this things which are done with sql like
>   that as you know
>
>   select site, count(site) from log_table
>   group by site
>
>   Mhy solution is: A second process which insert logs to table (temporary
>   table ) and aftr inserting do some quesries on this temporarty table and
> get
>  results to main statistics tablse which has tables as which site is most
>   visited table . This temporary table updates related statistics table .
>
>
>   I have need a recommendation about the problem? Is there any solution on
>   lucene(get most ranked kind query searching is exist or not and if yes is
> it
>  good for performance)
>   If there is no a solution in lucene what will you use for this situation?
> : Thanks for your help.
>
>
> : ilkay POLAT
>
> : Research &Development Software Engineer TURKEY
>



-- 
---
ilkay POLAT

Research &Development Software Engineer TURKEY