You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@nutch.apache.org by Marko Bauhardt <mb...@media-style.com> on 2005/11/08 13:40:02 UTC

index folder structure

Hello all,
i use nutch from the mapred branch. I have a little problem with the  
folder structure from the generated index. If i index a segment in a  
folder with name "indexes" the folder structure is
$NUTCH_HOME/indexes/part-00000
$NUTCH_HOME/indexes/part-00001
$NUTCH_HOME/indexes/part-00002
....

So, if i index my next segment in this folder, the old index will be  
overwritten. In a result of that i index every segment to indexes/ 
SEGMENT_NAME. Now my folder structure is:
$NUTCH_HOME/indexes/SEGMENT_NAME/part-00000
$NUTCH_HOME/indexes/SEGMENT_NAME/part-00001
$NUTCH_HOME/indexes/SEGMENT_NAME/part-00002
$NUTCH_HOME/indexes/OTHER_SEGMENT_NAME/part-00000
$NUTCH_HOME/indexes/OTHER_SEGMENT_NAME/part-00001
$NUTCH_HOME/indexes/OTHER_SEGMENT_NAME/part-00002
...

With this folder structure the searcher found no index. I look at the  
code and i think the NutchBean (method init) looks only in the first  
level depth to find the index.done file.
It would be nice, if the NutchBean search the index.done file in a  
level depth 2.

An other solution is the merging from my generated index. But i think  
this is only a workaround.

Or has anybody an other solution.

Thanks and bye, Marko