My team is new to swish-e. We're about to turn a bunch of people lose on new
content we've indexed. A little curious about the merging of indexes. It is
not well documented in things I've read so far.
I have one group of 19,000 HTML files - other indexes are smaller. I am
getting a document index rate via http method of 44 docs per minute. This is
not fast enough. If I divide the work up to 4 indexing processes (I have a
4way HPUX server), is the merge index command the thing to merge the four
indexes into one?
Anyone want to share experiences with large document collections and indexing
performance ideas? If I index the 19000 docs I have now, what if I have
a script to index only changed/new docs - can I merge back into the existing
index? Will it handle a changed doc and new docs?
Received on Thu Feb 17 02:36:04 2000