Xapian discuss - Jan 2006 - issues with accessing an index

I posted this question around the time you were having problems with your 
mail servers and didn't get any response, so I am trying again. I'm
having
the following problems on two of my xapian databases. The others dbs work 
fine

1) When I try to use quartzompact on a database it stops with the following:

"termlist ...quartzcompact: The revision being read has been discarded -
you
should call Xapian::Database::reopen() and retry the operation"

2) When I try to query another database it stops midway with

"terminate called after throwing an instance of
'Xapian::DatabaseModifiedError'
Aborted"

Please let me know if you have any idea what this means and what I can do to 
get around it.
Thanks
salem

On Tue, Jan 24, 2006 at 05:59:42PM +0000, Salem Berhanu wrote:
> 1) When I try to use quartzompact on a database it stops with the
following:
> 
> "termlist ...quartzcompact: The revision being read has been discarded
- you
> should call Xapian::Database::reopen() and retry the operation"
You're indexing into the database while trying to run quartzcompact on
it.
> 2) When I try to query another database it stops midway with
> 
> "terminate called after throwing an instance of
> 'Xapian::DatabaseModifiedError'
> Aborted"
You're indexing into the database quite aggressively while running
queries against it.

There are a couple of solutions, one being to index to one db then
copy to the one you index (perhaps via a quartzcompact mechanism).

In the future, flint (I believe) will get round this problem more
neatly.

J

-- 
/--------------------------------------------------------------------------\
  James Aylett                                                  xapian.org
  james@tartarus.org                               uncertaintydivision.org

On Thu, Feb 02, 2006 at 03:00:54PM +0000, Salem Berhanu wrote:
> I am not storing anything in the document data other than the
> ids. Eventually I will link to an external database but I will do it
> in ranges of not more than 50 at a time. However, I will need the
> initial compete ids to combine with results from other xapian
> dbs. This is because each document is broken up into chunks (since
> the information can be logically divided) and indexed in separate
> dbs. (eg. there is a title db, a description db ... ) I want to be
> able to combine the results across these dbs using boolean
> expressions (since I am assuming there isn't a built in way of doing
> this).
I'm sorry, there's not much point in trying to answer questions like
this without understand what you're actually trying to achieve. I
don't know why you're using several different databases, for
instance.

Let's go back to the beginning. What sort of performance issues are
you actually seeing? Have you investigated performance using the
standard performance tools for your platform? For instance, are you
running out of file buffer cache at the point that you start accessing
the document data? Other people have handled larger data sets that
you, so it may be that your hardware or configuration isn't quite
right for what you're trying to do.

(And please keep the discussion on list where others can help you as
well :-)

Cheers,
James

-- 
/--------------------------------------------------------------------------\
  James Aylett                                                  xapian.org
  james@tartarus.org                               uncertaintydivision.org