Skip to main content.
home | support | download

Back to List Archive

Re: [swish-e] incremental indexes

From: Peter Karman <peter(at)>
Date: Sun Mar 23 2008 - 01:31:59 GMT
Brandon Shalton wrote on 3/22/08 6:56 PM:
> hello all,
> i have started to use incremental indexing option because i wanted to be 
> able to batch add to the index.  all was going great until i hit the 
> 2gig mark on the index file.
> i have started another index file, so i will let that run until it 
> reaches the 2gig file limit.

What version are you using? Recent SVN snapshots have automatic LARGEFILE option 
detection at configure time. Otherwise, you likely need to specify that to the 
configure script.

> in looking at the files, i see the .idx and .wdata files at 2gig. 
> zipping them up to bring them back from a remote server would still make 
> for a huge file.
> is there a way to merge the incremental index files to 1 index file such 
> that its not in incremental mode?

no. the incremental (btree) version has a totally different set of backend code 
than the native non-incremental version.

You might try the 2.6 svn branch. It uses Berkeley DB backend.

Peter Karman  .  .  peter(at)
Users mailing list
Received on Sat Mar 22 21:32:10 2008