Closed GoogleCodeExporter closed 8 years ago
I have the same problem.
Original comment by bill.ms....@gmail.com
on 26 May 2010 at 8:25
I had the same problem. After the index creation I had many different files in
my index folder (*.fdt, *.fdx, *.tvf, *.tvx, etc.). Whet I tried to open the
index, there always came an error.
Then I created the index once more AND did "myIndexWriter.close()" at the end
of the indexing process. After that I had ony 3 files in the index folder
(_0.cfs, semgents.gen, segments_2) and I could open the index with Luke without
any errors.
Hope that helps you (or others with this issue?).
Original comment by karolina...@googlemail.com
on 18 Jan 2011 at 1:20
This behavior of Lucene and Luke is as designed. It's there to prevent
accidental index corruption if there is another process that is still writing
to the index. Therefore it makes sense that it should require a write access
and it should be a conscious decision that you really want to modify an
existing index (unlocking is a modification).
Original comment by sig...@gmail.com
on 27 Apr 2011 at 10:19
Huh? Why does it make sense that it should require write access? I came across
this when trying to read an index that was currently being written by another
process. I know you shouldn't have more than one concurrent writer, which is
why I selected "Open in Read-Only mode". Seems like Luke should allow
read-only access to a locked index, since that's what Lucene is designed to
support.
Original comment by martin.m...@gmail.com
on 6 Sep 2011 at 4:07
Original issue reported on code.google.com by
dcli...@gmail.com
on 7 Apr 2010 at 12:13