You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@jena.apache.org by "Paul Gearon (JIRA)" <ji...@apache.org> on 2012/06/21 03:00:57 UTC
[jira] [Created] (JENA-266) MINUS operations too slow
Paul Gearon created JENA-266:
--------------------------------
Summary: MINUS operations too slow
Key: JENA-266
URL: https://issues.apache.org/jira/browse/JENA-266
Project: Apache Jena
Issue Type: Improvement
Components: ARQ
Reporter: Paul Gearon
MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398451#comment-13398451 ]
Paul Gearon commented on JENA-266:
----------------------------------
Yes, IndexTable could certainly use some. I'll add them in shortly.
Adding the test scripts to the SPARQL 1.1 test suite now.
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399428#comment-13399428 ]
Paul Gearon commented on JENA-266:
----------------------------------
The main part of the patch was TestIndexTable, but this hasn't been added. Was that intentional?
src/test/java/com/hp/hpl/jena/sparql/engine/index/TestIndexTable.java
(src/test/java/com/hp/hpl/jena/sparql/engine/index is a new directory)
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399110#comment-13399110 ]
Paul Gearon edited comment on JENA-266 at 6/22/12 4:45 AM:
-----------------------------------------------------------
Attached a new patch to test HashIndexTable, LinearIndex and IndexFactory.
Patch also added some comments and hubris to each of the index classes to explain their purpose. Fixed a copy/paste typo in QueryIter2
was (Author: quoll):
Tests for HashIndexTable, LinearIndex and IndexFactory.
Added some comments and hubris to each of the index classes to explain their purpose.
Fixed a copy/paste typo in QueryIter2
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398543#comment-13398543 ]
Hudson commented on JENA-266:
-----------------------------
Integrated in Jena_ARQ #666 (See [https://builds.apache.org/job/Jena_ARQ/666/])
Applied Paul Gearon's patch for MINUS evaluation improvements (JENA-266) (Revision 1352595)
Result = SUCCESS
rvesse :
Files :
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/algebra/index
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/algebra/index/HashIndexTable.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/algebra/index/IndexFactory.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/algebra/index/IndexTable.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/algebra/index/LinearIndex.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/iterator/QueryIterMinus.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/main/OpExecutor.java
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398588#comment-13398588 ]
Andy Seaborne commented on JENA-266:
------------------------------------
Yes, a bit spooky isn't? Jenkins seems to look at commit logs and poke JIRA based on any JIRA label.
It is only the module build done, without deployment (left from when we built things separately) - the whole of apache-jena is always and only built over night. If a module changes an external interface, a manual "mvn clean deploy" is needed.
Any Jena commiter can go into Jenkins and request a build immediately.
Jena is not a very large build (TDB is bit slow because of disk-based tests). We could change to a a whole-system build after a commit.
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398225#comment-13398225 ]
Rob Vesse commented on JENA-266:
--------------------------------
Patch looks good (have only eyeballed it so far) - will go ahead and apply and test tomorrow morning
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398600#comment-13398600 ]
Rob Vesse commented on JENA-266:
--------------------------------
A whole system build might be appropriate in case someone makes a change in one module and only builds and tests that module so doesn't see they break something over in another module that depends on the affected module
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Andy Seaborne resolved JENA-266.
--------------------------------
Resolution: Fixed
Fix Version/s: ARQ 2.9.2
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Fix For: ARQ 2.9.2
>
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398572#comment-13398572 ]
Rob Vesse commented on JENA-266:
--------------------------------
Thanks Andy, btw I notice applying this patch created ARQ build 666
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398610#comment-13398610 ]
Andy Seaborne commented on JENA-266:
------------------------------------
(we are off topic!)
That gets caught by the overnight build (which is the release build BTW).
It's two cycles - inner one is per module (supposed quick feedback but of course we all remember to run the per-module tests before commit don't we :-). The outer cycle is the dev buidl of everything from scratch.
Unfortunately, the dev build then causes all the modules dependencies to change and they all rebuild (at a time when the system is quiet-ish).
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Paul Gearon updated JENA-266:
-----------------------------
Attachment: minus-tests.patch
Just realized that both patches had the same filename. Removed the patch containing tests and re-submitted.
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Closed] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Andy Seaborne closed JENA-266.
------------------------------
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Fix For: ARQ 2.9.2
>
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Paul Gearon updated JENA-266:
-----------------------------
Attachment: (was: arq.patch)
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398423#comment-13398423 ]
Andy Seaborne commented on JENA-266:
------------------------------------
Looks good. (I applied the patch to a scratch copy and nothing seems to have broken!)
What would be the best way to test this?
It would be nice to have some code @Tests of IndexTable, and some script tests (we can get the latter from SPARQL 1.1 test suite).
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Paul Gearon updated JENA-266:
-----------------------------
Attachment: arq.patch
Tests for HashIndexTable, LinearIndex and IndexFactory.
Added some comments and hubris to each of the index classes to explain their purpose.
Fixed a copy/paste typo in QueryIter2
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Andy Seaborne (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398534#comment-13398534 ]
Andy Seaborne commented on JENA-266:
------------------------------------
I moved the package "index" to be under "engine" -- the "algebra" package is the static structures for the algebra, and the execution under engine (roughly - it's not a completely clean split).
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399419#comment-13399419 ]
Rob Vesse commented on JENA-266:
--------------------------------
New patch has been applied, Hudson should comment shortly
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399437#comment-13399437 ]
Hudson commented on JENA-266:
-----------------------------
Integrated in Jena_ARQ #669 (See [https://builds.apache.org/job/Jena_ARQ/669/])
Applied Paul Gearon's further patch for JENA-266 to add test cases for new data structures introduced to speed up MINUS evaluation (Revision 1352959)
Result = SUCCESS
rvesse :
Files :
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/index/HashIndexTable.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/index/IndexFactory.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/index/IndexTable.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/index/LinearIndex.java
* /jena/trunk/jena-arq/src/main/java/com/hp/hpl/jena/sparql/engine/iterator/QueryIter2.java
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13399435#comment-13399435 ]
Rob Vesse commented on JENA-266:
--------------------------------
Sorry my mistake, am working on a few different things for ARQ atm some of which are also unversioned so had Show Unversioned Files off in the commit dialogue to limit to just the things I wanted committed and missed this. Should now be applied
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch, minus-tests.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398523#comment-13398523 ]
Rob Vesse commented on JENA-266:
--------------------------------
Applied in Trunk now so it can get into snapshots and allow people to test this out. Some more detailed unit tests for this would be nice so I will leave the issue open for the time being until we have those in place
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (JENA-266) MINUS operations too slow
Posted by "Paul Gearon (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Paul Gearon updated JENA-266:
-----------------------------
Attachment: arq.patch
Patch against jena-arq
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (JENA-266) MINUS operations too slow
Posted by "Rob Vesse (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/JENA-266?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Rob Vesse reassigned JENA-266:
------------------------------
Assignee: Rob Vesse
> MINUS operations too slow
> -------------------------
>
> Key: JENA-266
> URL: https://issues.apache.org/jira/browse/JENA-266
> Project: Apache Jena
> Issue Type: Improvement
> Components: ARQ
> Reporter: Paul Gearon
> Assignee: Rob Vesse
> Attachments: arq.patch
>
>
> MINUS operations use an N^2 algorithm when evaluating. Despite the data being placed on the heap, an operating removing ~100k bindings from ~500k bindings takes >15 minutes on desktop systems (15 minutes being the time at which the process is killed. It may take significantly longer).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira