lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hoss Man (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (SOLR-5944) Support updates of numeric DocValues
Date Wed, 25 May 2016 22:53:13 GMT

     [ https://issues.apache.org/jira/browse/SOLR-5944?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Hoss Man updated SOLR-5944:
---------------------------
    Attachment: SOLR-5944.patch

Ugh, testReplay4 in my last patch was a waste of time -- somehoe i screwed something up with
copy/paste.

The point was to try and check if using "set" operations had any different effect on the test
compared to using "inc" operations (like in testReplay3)

This patch fixes that, so that both tests demonstrate the problem with the seed ishan mentioned
(and aren't identical) ... but more interesting then that is the new testReplay5, testReplay6
& testReplay7 (still using hte same seed)...

* testReplay5 - still uses "inc" for doc id=0, but uses "set" for every other doc in the index
** this currently fails with an NPE in {{AtomicUpdateDocumentMerger.doInPlaceUpdateMerge(AtomicUpdateDocumentMerger.java:283)}}
* testReplay6 was a quick hack to work around the NPE in testReplay5 - ensuring that the first
time a doc is added, a regular "add" is done, but everytime after that "set" is used (except
for id=0 where "inc" is still used)
** this test currently passes -- even though it should be effectively the same as testReplay3
and testReplay4
* testReplay7 is the same as testReplay6 but (like testReplya4) does all the id=0 updates
using "set" as well just to sanity check no diff in behavior
** this also passes

...so based on the fact that testReplay6/7 pass, but testReplay3/4 fail, any ishan's observations
about the setMaxBufferedDocs my suspicion is that the underlying problem probably has something
to do with how frequently docs are flushed and trying to read back "updated" DV values when
other docs were added with regular DV values i nte hsame field just before/after the DV update.

(looking at TestNumericDocValuesUpdates i don't see many tests that mix calls to addDocument
with calls to updateNumericDocValue -- except testSegmentMerges & testManyReopensAndFields
which update the DV of *every* document added up to that point)

> Support updates of numeric DocValues
> ------------------------------------
>
>                 Key: SOLR-5944
>                 URL: https://issues.apache.org/jira/browse/SOLR-5944
>             Project: Solr
>          Issue Type: New Feature
>            Reporter: Ishan Chattopadhyaya
>            Assignee: Shalin Shekhar Mangar
>         Attachments: DUP.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch,
SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch,
SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch,
SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch,
SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch, SOLR-5944.patch,
SOLR-5944.patch, TestStressInPlaceUpdates.eb044ac71.beast-167-failure.stdout.txt, TestStressInPlaceUpdates.eb044ac71.beast-587-failure.stdout.txt,
TestStressInPlaceUpdates.eb044ac71.failures.tar.gz, hoss.62D328FA1DEA57FD.fail.txt, hoss.62D328FA1DEA57FD.fail2.txt,
hoss.62D328FA1DEA57FD.fail3.txt, hoss.D768DD9443A98DC.fail.txt, hoss.D768DD9443A98DC.pass.txt
>
>
> LUCENE-5189 introduced support for updates to numeric docvalues. It would be really nice
to have Solr support this.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message