[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498652
]
Michael Busch commented on LUCENE-888:
--
> At 1 KB (current Lucene trunk) it takes 622 sec to build the index; if
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498696
]
Michael McCandless commented on LUCENE-888:
---
> > I'm guessing we should leave BufferedIndexInput's default
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498714
]
Michael Busch commented on LUCENE-888:
--
> I wonder if we should just add a ctor to BufferedIndexInput that takes
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498739
]
Michael McCandless commented on LUCENE-888:
---
> > I wonder if we should just add a ctor to BufferedIndexInpu
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498757
]
Michael Busch commented on LUCENE-888:
--
> Actually, it does have a default public constructor right? Ie if we ad
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498776
]
Michael McCandless commented on LUCENE-888:
---
> > I plan to add "private int bufferSize" to BufferedIndexInp
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498780
]
Michael Busch commented on LUCENE-888:
--
> OK I will do this. Actually, I think we should also allow making the
>
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498784
]
Michael McCandless commented on LUCENE-888:
---
> > OK I will do this. Actually, I think we should also allow
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498787
]
Marvin Humphrey commented on LUCENE-888:
I would like to know why these gains are appearing, and how specific
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498793
]
Michael Busch commented on LUCENE-888:
--
> I'm actually coding it up now. Why don't you commit LUCENE-430
> sooni
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499018
]
Michael McCandless commented on LUCENE-888:
---
OK I ran two sets of tests. First is only on Mac OS X to see
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499020
]
Michael McCandless commented on LUCENE-888:
---
> I would like to know why these gains are appearing, and how
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499044
]
John Haxby commented on LUCENE-888:
---
> Net/net it's between 10-18% performance gain overall. It is
> interesting th
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499182
]
Michael Busch commented on LUCENE-888:
--
Mike,
I tested and reviewed your patch. It looks good and all tests pas
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499199
]
Michael McCandless commented on LUCENE-888:
---
> I tested and reviewed your patch. It looks good and all test
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499206
]
Michael Busch commented on LUCENE-888:
--
> I'm a little nervous about that: I don't know the impact it will have
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499209
]
Michael McCandless commented on LUCENE-888:
---
> > I'm a little nervous about that: I don't know the impact i
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499212
]
Doug Cutting commented on LUCENE-888:
-
> then we don't save IO by limiting the buffer size to 1 KB
I'm confused
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499214
]
robert engels commented on LUCENE-888:
--
I think the important consideration is how expensive is the system call.
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499221
]
Michael Busch commented on LUCENE-888:
--
> So, back to your statement, a 1k buffer doesn't save
> physical i/o,
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499223
]
Michael Busch commented on LUCENE-888:
--
Mike,
another thing I just noticed is your new testcase doesn't remove
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499224
]
Eks Dev commented on LUCENE-888:
we did some time ago a few tests and simply concluded, it boils down to what
Doug s
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499225
]
Michael McCandless commented on LUCENE-888:
---
> another thing I just noticed is your new testcase doesn't re
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499226
]
Michael McCandless commented on LUCENE-888:
---
> we did some time ago a few tests and simply concluded, it b
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499238
]
Michael Busch commented on LUCENE-888:
--
Take2 looks good. +1 for committing.
> Improve indexing performance by
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499280
]
Marvin Humphrey commented on LUCENE-888:
I have some auxiliary data points to report after experimenting with
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499299
]
Michael McCandless commented on LUCENE-888:
---
Marvin, it's odd that you see no gains when coming straight fr
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499307
]
Marvin Humphrey commented on LUCENE-888:
> Also, how much "merging" is actually done in your test / KS? How
[
https://issues.apache.org/jira/browse/LUCENE-888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12499548
]
Michael McCandless commented on LUCENE-888:
---
I re-ran the "second test" above, but this time with compound
Sorry, have not noticed LUCENE-893... it was Searching, indexing was slightly
above 20%
Interesting! Do you remember if your 5-8% gain was for searching or
indexing? This issue is focusing on buffer size impact on indexing
performance and LUCENE-893 is focusing on search performance.
30 matches
Mail list logo