Well, this 'replacement' of the ThreadLocal does not solve the initial
problem. As there's always at least one ThreadLocal which binds the object
loaded by the web-app to the Thread which is _not_ loaded by the web-app,
the classloader never may be unloaded.
You are right, this is not the 'leak'
On Jul 11, 2008, at 1:13 PM, xin liu wrote:
I have individual index files for Audio, Image and PDF files. We
build common meta fields for these different data types. When I
search for a string, I want the search to return mixed search
results from these 3 different index based on relevancy.
11 jul 2008 kl. 15.28 skrev jnance:
The TermFrequencyVector works perfectly for normal query strings.
But if I
add a wild card (*) onto words to search for different forms of the
word I
get an ArrayIndexOutOfBoundsException because the index is -1. Why
does this
happen? And is there anyw
Hi,
I have individual index files for Audio, Image and PDF files. We build common
meta fields for these different data types. When I search for a string, I want
the search to return mixed search results from these 3 different index based on
relevancy. So I use ParallelMultiSearcher class to do
The TermFrequencyVector works perfectly for normal query strings. But if I
add a wild card (*) onto words to search for different forms of the word I
get an ArrayIndexOutOfBoundsException because the index is -1. Why does this
happen? And is there anyway to avoid it?
Thanks,
James
jnance wrot
OK, sounds good. Fall will be here before you know it!
Mike
Christopher Kolstad wrote:
The only way to make this work with svn is if you can have svn
perform a
switch without doing any removal, then restart your IndexSearcher,
then do a
normal svn switch to remove the now unused files.
>
> The only way to make this work with svn is if you can have svn perform a
> switch without doing any removal, then restart your IndexSearcher, then do a
> normal svn switch to remove the now unused files. Does svn have an option
> to "switch but don't remove any removed files"? Because IndexSe
After discussing this on java-dev:
http://mail-archives.apache.org/mod_mbox/lucene-java-dev/200807.mbox/[EMAIL
PROTECTED]
it seems that this is not in fact a leak but rather a delayed GC
issue. The objects are eventually freed, on Sun 1.4, 1.5 and 1.6.
When a ThreadLocal instance beco
OK, got it.
The only way to make this work with svn is if you can have svn perform
a switch without doing any removal, then restart your IndexSearcher,
then do a normal svn switch to remove the now unused files. Does svn
have an option to "switch but don't remove any removed files"?
Bec
The deleted docs are actually stored separately, per segment, into
files named _X_N.del, where X is the segment name and N is a
generation count (keeps increasing by 1 every time new deletes are
committed to that segment).
Normal segment merging will also collapse the deletes in those
s
I think the FieldCache API fits for this...
Its purpose is to do a one-time caching of a particular field's
values, rendered as a straight array keyed by document ID. So
building that cache initially takes time and memory, but then getting
the value for a particular document is extremely
You're welcome!
And, feel free to go vote for that issue ... Jira's voting system lets
us gauge importance of features like this over time :)
Mike
Aditi Goyal wrote:
Thanks Mike for your valuable time.
Regards,
Aditi
On Thu, Jul 10, 2008 at 5:36 PM, Michael McCandless <
[EMAIL PROTECTED
Hi.
First, thanks for the reply.
Why does SubversionUpdate require shutting down the IndexSearcher? What
> goes wrong?
>
SubversionUpdate requires shutting down the IndexSearcher in our current
implementation because the old index files are deleted in the tag we're
switching to. Sorry, just rea
13 matches
Mail list logo
