You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@couchdb.apache.org by "Adam Kocoloski (JIRA)" <ji...@apache.org> on 2010/04/20 15:54:49 UTC

[jira] Updated: (COUCHDB-746) More speed improvements for default view collation

     [ https://issues.apache.org/jira/browse/COUCHDB-746?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Adam Kocoloski updated COUCHDB-746:
-----------------------------------

    Attachment: less_json.patch

> More speed improvements for default view collation
> --------------------------------------------------
>
>                 Key: COUCHDB-746
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-746
>             Project: CouchDB
>          Issue Type: Improvement
>          Components: Database Core
>    Affects Versions: 0.11
>            Reporter: Adam Kocoloski
>             Fix For: 1.1
>
>         Attachments: less_json.patch
>
>
> We left Paul's less_json.patch hanging in COUCHDB-495.  I also noticed during an eprof run that all the function calls to type_sort, less_same_type, etc. seemed to be taking a non-trivial amount of time.  Here's a patch to try to address both those issues.
> I ran perf.py from COUCHDB-495 and got the following MegaView numbers
> trunk
> 6.12
> 5.77
> 5.99
> 6.07
> 5.84
> patch
> 5.73
> 4.94
> 4.75
> 5.00
> 5.28
> Taking the minimum of each it looks like an 18% boost in indexing speed.  The SimpleView numbers were too close to differentiate.
> There is one small change which I think is harmless -- I removed the code to handle arbitrary tuples (as opposed to proplists).  I don't see how a JSON decoder could produce tuples which are not proplists.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.