lucene-java-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael McCandless <>
Subject Re: Upgrading from v2.2.0 to v2.3.2
Date Wed, 27 Aug 2008 09:50:10 GMT

Mark Lassau wrote:

> Mike,
> Thanks for the prompt response.
> Michael McCandless wrote:
>> Mark Lassau wrote:
>>> I am a developer on the JIRA Issue tracker, and we are considering  
>>> upgrading our Lucene version from v2.2.0 to v2.3.2.
>>> I have been charged with doing the risk analysis, and project work.
>>> I have read the change lists, and the bugs reported on the Lucene  
>>> Issue Tracker (JIRA of course ;), and some of the archived  
>>> messages on the Lucene lists.
>> We love Jira -- thanks!
> ... and we love Lucene :)

Glad to hear it!

> <snip>
>> I would think LUCENE-1044 (not corrupting search index if OS or  
>> machine crashes) is important for Atlassian.
> LUCENE-1044 is only marked as fixed in v2.4
> I assumed that the possible corruption on hard shutdown existed in  
> v2.2.0 (and earlier), so we are no worse off by upgrading to v2.3.2.
> Is that the case?

That's right.

>> The CHANGES.txt should be a reliable list of any non-backwards- 
>> compatible changes and how to revert to the old behavior (and if it  
>> isn't we'd love to hear about that!).  EG merging is now done  
>> concurrently by default, flushing is by RAM usage not doc count,  
>> merge selection is by size-in-bytes of each segment (not size-in- 
>> doc-count).
> Thanks, I read those. Good job on providing a very useful change  
> summary.
> However, I guess I was also interested in any reports "from the  
> field". Sometimes there can be unexpected consequences.
> eg, I read somewhere of one user having multi-threaded issues  
> because of a change from Vector to ArrayList in some implementation.

OK gotchya.

I think that particular change was to Document (in how it stores  
fields internally); there was a case where someone was actually adding  
fields to a document with multiple threads and the switch to ArrayList  
broke that code.  It's tricky because Document never declared that its  
APIs were thread safe.


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message