directory-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Kiran Ayyagari <>
Subject Re: Problem with indexes...
Date Mon, 12 May 2014 16:46:24 GMT
On Mon, May 12, 2014 at 9:08 PM, Mark D <> wrote:

> Kiran, I don't have a reliable test.  I created a new snapshot build from
> trunk and
> I noticed the following in my logs.  I don't recall seeing these before.
>  I looked
> at previous logs and didn't find examples but I didn't have a really good
> set to compare with.
> [11:10:22] INFO
> []
> - building the index for attribute type attributetype (
> 0.9.2342.19200300.100.1.1 NAME ( 'uid' 'userid' )
> Restarts appear to be working without loosing search, good news.
> I would like to figure out what is causing this ?  Guessing I have some
> index I auto created that is not defined / used in the partition?
> this happens when a new index was configured _after_ the partition was
loaded/added with some data
and the server detects this during next restart and builds the index.

If you are curious to know why the search started working now:

 Earlier the index database file was created (after configuring and
restarting the server) but the existing
 data was not indexed into this file, so the index scan on say 'uid'
attribute returned zero results (note that
 a full table scan is avoided cause the index file is present).

 now the old/existing data gets indexed and the index file is appropriately
filled with that data resulting in
 correct search results.

[11:10:22] WARN
> []
> - Failed to delete unused index file
> C:\Users\mdeb\workspace-jimbra\Jaimbra\work\apacheds\jimbra\partitions\jimbra\
> > this is not related to the above issue (cause once data is in the index
> it
> will survive restarts),
> do you have a test case that I can run to reproduce?
> Thanks, Makr.

Kiran Ayyagari

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message