Sharad Agarwal wrote:
I am a newbie in lucene space. and trying to understand lucene search
result caching; facing with a wierd issue.
After creating the IndexReader from a file system directory, I
rename/remove the index directory; but still I am able to search the
index and able to get the documents from Hits. This is making me beleive
that there is some level of index loading in IndexReader itself. Does it
load the full index or what all things it loads?
If you're running on Linux, your process has open file descriptors on the
index files. So, even though you've rm'd them, they're not really gone as far
as the kernel is concerned until those file descriptors are closed (the space
isn't reclaimed until then either).
I don't know if Windows acts the same way.
I'm curious: Why are you removing the index if your application is still
using it?
--MDC
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
