lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Muir (Commented) (JIRA)" <>
Subject [jira] [Commented] (LUCENE-3509) Add settings to IWC to optimize IDV indices for CPU or RAM respectivly
Date Sun, 23 Oct 2011 21:34:32 GMT


Robert Muir commented on LUCENE-3509:

well the same applies to having an IndexDocValuesConfig.

If my implementation doesn't use packed integers, it doesnt make a lot of sense?

So I think it would be best if this was only a parameter to this particular implementation...
otherwise we end out with a ton of options across all possible implementations, which won't
make a lot of sense at all.

I think at the end of the day, we should imagine having tons of implementations for the different
index parts (look how many term dictionaries we have), and let impls have their own parameters.

If someone wants to toggle these crazy parameters, thats something they can do in their codec
definition. For solr users etc, thats something they can tweak in the xml on a per-implementation
basis just like when they provide crazy parameters to various tokenfilters in their analysis

> Add settings to IWC to optimize IDV indices for CPU or RAM respectivly
> ----------------------------------------------------------------------
>                 Key: LUCENE-3509
>                 URL:
>             Project: Lucene - Java
>          Issue Type: Improvement
>          Components: core/index
>    Affects Versions: 4.0
>            Reporter: Simon Willnauer
>            Priority: Minor
>             Fix For: 4.0
>         Attachments: LUCENE-3509.patch
> spinnoff from LUCENE-3496 - we are seeing much better performance if required bits for
PackedInts are rounded up to a 8/16/32/64. We should add this option to IWC and default to
round up ie. more RAM & faster lookups.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


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

View raw message