Hi,

The indexing followed by the merge of index files is
all in a single perl script. So, once the index files
are generated the merge command is executed via the
backtik operator.

This works fine for a smaller set/size of index files.
However, the merge process ends abrubtly when trying
to merge larger index files. 

Since the merge stopped I have now again executed the
merge command seprately redirecting the output and
error messages into seperate files.

Here are the last bit from the output file

Processing words in index '/swishe/new_rose.index':
20985685 words
Removed   2051 words no longer present in docs for
index '/swishe/new_rose.index'

Writing main index...
Sorting words ...
Sorting 20983634 words alphabetically
Writing header ...
Writing index entries ...
  Writing word text: ...
  Writing word text:  10%
  Writing word text:  20%
  Writing word text:  30%
  Writing word text:  40%
  Writing word text:  50%
  Writing word text:  60%
  Writing word text:  70%
  Writing word text:  80%
  Writing word text:  90%
  Writing word text: 100%
  Writing word text: Complete
  Writing word hash: ...err: Ran out of memory (could
not allocate 251803608 more bytes)!

<<
A memory problem .. :)
The server I am running this on has 2 Gb RAM and is a
dual processor.


Guess this info should be of some help.

Ashim

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com