[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-19 Thread Robert Muir (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713360#comment-13713360
 ] 

Robert Muir commented on LUCENE-5121:
-

{quote}
Well the fields that we use for docValues are defined as below. Whats is 
non-standard here?
{quote}

The wiki page is quite clear about this: 
http://wiki.apache.org/solr/DocValues#Specifying_a_different_Codec_implementation

{quote}
Where as my solrconfig.xml file i had to add this line to make it work. (cause 
we upgraded from 3.5 to 4.2.1)

 
codecFactory class=solr.SchemaCodecFactory/
{quote}

Its too bad you didn't also include the comment that goes with that line in 
solrconfig.xml, which also warns about this!

{noformat}
  !-- The CodecFactory for defining the format of the inverted index.
   The default implementation is SchemaCodecFactory, which is the official 
Lucene
   index format, but hooks into the schema to provide per-field 
customization of
   the postings lists and per-document values in the fieldType element
   (postingsFormat/docValuesFormat). Note that most of the alternative 
implementations
   are experimental, so if you choose to customize the index format, its a 
good
   idea to convert back to the official format e.g. via 
IndexWriter.addIndexes(IndexReader)
   before upgrading to a newer version to avoid unnecessary reindexing.
  --
{noformat}

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-19 Thread Yonik Seeley (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713616#comment-13713616
 ] 

Yonik Seeley commented on LUCENE-5121:
--

I guess to the casual observer / user, it's not clear what is covered by the 
default codec and what is considered an experimental codec.

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-19 Thread Aditya (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713712#comment-13713712
 ] 

Aditya commented on LUCENE-5121:


So this concludes that for non-default codec we need to re-index on upgrade. 

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-19 Thread Robert Muir (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713795#comment-13713795
 ] 

Robert Muir commented on LUCENE-5121:
-

No, this is unnecessary. Again, please read the documentation (its all been 
pasted verbatim here on this issue)

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Robert Muir (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713044#comment-13713044
 ] 

Robert Muir commented on LUCENE-5121:
-

We only provide back compat for the default codec.

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Robert Muir (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713050#comment-13713050
 ] 

Robert Muir commented on LUCENE-5121:
-

Again not a bug...

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Hoss Man (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713053#comment-13713053
 ] 

Hoss Man commented on LUCENE-5121:
--

bq. We only provide back compat for the default codec.

WTF?  Seriously?

1) is this documented anywhere?

2) I thought a major point of having multiple codecs and configurable posting 
formats and docvalues formats was that the old ones could be left alone for 
backcompat support, and entirely new ones, with new names, could be created 
if/when there were format improvements to be made?



 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Hoss Man (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713055#comment-13713055
 ] 

Hoss Man commented on LUCENE-5121:
--

FWIW, the original stack trace reported by the user on IRC...

{noformat}
Caused by: org.apache.lucene.index.CorruptIndexException: invalid type: 65, 
resource=MMapIndexInput(path=/storage/solrdata/index/_fd_Disk_0.dvdm)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesProducer.readFields(DiskDocValuesProducer.java:159)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesProducer.init(DiskDocValuesProducer.java:72)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesFormat.fieldsProducer(DiskDocValuesFormat.java:49)
at 
org.apache.lucene.codecs.perfield.PerFieldDocValuesFormat$FieldsReader.init(PerFieldDocValuesFormat.java:213)
at 
org.apache.lucene.codecs.perfield.PerFieldDocValuesFormat.fieldsProducer(PerFieldDocValuesFormat.java:282)
at 
org.apache.lucene.index.SegmentCoreReaders.init(SegmentCoreReaders.java:134)
at org.apache.lucene.index.SegmentReader.init(SegmentReader.java:56)
at 
org.apache.lucene.index.StandardDirectoryReader$1.doBody(StandardDirectoryReader.java:62)
at 
org.apache.lucene.index.SegmentInfos$FindSegmentsFile.run(SegmentInfos.java:812)
at 
org.apache.lucene.index.StandardDirectoryReader.open(StandardDirectoryReader.java:52)
at org.apache.lucene.index.DirectoryReader.open(DirectoryReader.java:88)
at 
org.apache.solr.core.StandardIndexReaderFactory.newReader(StandardIndexReaderFactory.java:34)
at 
org.apache.solr.search.SolrIndexSearcher.getReader(SolrIndexSearcher.java:174)
... 18 more

{noformat}

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Robert Muir (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713070#comment-13713070
 ] 

Robert Muir commented on LUCENE-5121:
-

This is documented everywhere, see solrconfig.xml or the docvalues wiki page 
for example.

Having a codec API doesn't mean expanding our back compat requirements to 
custom/experimental formats. If you don't want to reindex, just merge to the 
default codec, then upgrade.


 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Mark Miller (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713103#comment-13713103
 ] 

Mark Miller commented on LUCENE-5121:
-

The wiki page says: 

Note that only the default implementation is supported by future version of 
Lucene: if you try an alternative format, you may need to switch back to the 
default and rewrite your index (e.g. forceMerge) before upgrading.

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Aditya (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713156#comment-13713156
 ] 

Aditya commented on LUCENE-5121:


Well the fields that we use for docValues are defined as below. Whats is 
non-standard here?
{code:xml} 
 fieldType name=dv_date class=solr.TrieDateField omitNorms=true 
precisionStep=6 docValuesFormat=Disk positionIncrementGap=0 /
 
 dynamicField name=lcontNumOfDownload.* type=dv_long indexed=true 
stored=true default=0  docValues=true/
{code}

Where as my solrconfig.xml file i had to add this line to make it work.
{code:xml} 
codecFactory class=solr.SchemaCodecFactory/
{code}

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when using CheckIndex -- 
 details to reproduce to follow in an attachment/comment

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

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



[jira] [Commented] (LUCENE-5121) 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals

2013-07-18 Thread Aditya (JIRA)

[ 
https://issues.apache.org/jira/browse/LUCENE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13713238#comment-13713238
 ] 

Aditya commented on LUCENE-5121:


Some additional information as Hoss Man asked on solr IRC after running 
CheckIndex with 4.4 and 4.2.1 

*With CheckIndex (4.4)*
{code}
[root@solrperfmv002 storage]# java -cp 
lucene-core-4.4.0.jar:lucene-codecs-4.4.0.jar 
org.apache.lucene.index.CheckIndex ./solrdata/index/

NOTE: testing will be more thorough if you run java with 
'-ea:org.apache.lucene...', so assertions are enabled

Opening index @ ./solrdata/index/

Segments file=segments_q numSegments=1 version=4.2.1 format= 
userData={commitTimeMSec=1374020371815}
  1 of 1: name=_fd docCount=24429799
codec=Lucene42
compound=false
numFiles=16
size (MB)=25,112.699
diagnostics = {timestamp=1374016842452, os=Linux, 
os.version=2.6.18-164.11.1.el5, mergeFactor=3, source=merge, 
lucene.version=4.2.1-SNAPSHOT 1490837M - greg - 2013-06-07 14:10:44, 
os.arch=amd64, mergeMaxNumSegments=1, java.version=1.6.0_45, java.vendor=Sun 
Microsystems Inc.}
no deletions
test: open reader.FAILED
WARNING: fixIndex() would remove reference to this segment; full exception:
org.apache.lucene.index.CorruptIndexException: invalid type: 65, 
resource=MMapIndexInput(path=/storage/solrdata/index/_fd_Disk_0.dvdm)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesProducer.readFields(DiskDocValuesProducer.java:159)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesProducer.init(DiskDocValuesProducer.java:72)
at 
org.apache.lucene.codecs.diskdv.DiskDocValuesFormat.fieldsProducer(DiskDocValuesFormat.java:49)
at 
org.apache.lucene.codecs.perfield.PerFieldDocValuesFormat$FieldsReader.init(PerFieldDocValuesFormat.java:213)
at 
org.apache.lucene.codecs.perfield.PerFieldDocValuesFormat.fieldsProducer(PerFieldDocValuesFormat.java:282)
at 
org.apache.lucene.index.SegmentCoreReaders.init(SegmentCoreReaders.java:134)
at org.apache.lucene.index.SegmentReader.init(SegmentReader.java:56)
at org.apache.lucene.index.CheckIndex.checkIndex(CheckIndex.java:543)
at org.apache.lucene.index.CheckIndex.main(CheckIndex.java:1854)

WARNING: 1 broken segments (containing 24429799 documents) detected
WARNING: would write new segments file, and 24429799 documents would be lost, 
if -fix were specified
{code}


*With CheckIndex (4.2.1)*
{code}
[root@solrperfmv002 storage]# java -d64 -Xmx16g -cp 
lucene-codecs-4.2.1-SNAPSHOT.jar:lucene-core-4.2.1-SNAPSHOT.jar 
org.apache.lucene.index.CheckIndex ./solrdata/index/

NOTE: testing will be more thorough if you run java with 
'-ea:org.apache.lucene...', so assertions are enabled

Opening index @ ./solrdata/index/

Segments file=segments_q numSegments=1 version=4.2.1 format= 
userData={commitTimeMSec=1374020371815}
  1 of 1: name=_fd docCount=24429799
codec=Lucene42
compound=false
numFiles=16
size (MB)=25,112.699
diagnostics = {timestamp=1374016842452, os=Linux, 
os.version=2.6.18-164.11.1.el5, mergeFactor=3, source=merge, 
lucene.version=4.2.1-SNAPSHOT 1490837M - greg - 2013-06-07 14:10:44, 
os.arch=amd64, mergeMaxNumSegments=1, java.version=1.6.0_45, java.vendor=Sun 
Microsystems Inc.}
no deletions
test: open reader.OK
test: fields..OK [12367 fields]
test: field norms.OK [515 fields]
test: terms, freq, prox...OK [139455881 terms; 2613497338 terms/docs pairs; 
80324 tokens]
test: stored fields...OK [2351712224 total field count; avg 96.264 
fields per doc]
test: term vectorsOK [0 total vector count; avg 0 term/freq vector 
fields per doc]
test: docvalues...OK [0 total doc count; 335 docvalues fields]

No problems were detected with this index.

{code}

 4.4 RC0 can not read indexes created with 4.2.1 using Disk based docvals
 

 Key: LUCENE-5121
 URL: https://issues.apache.org/jira/browse/LUCENE-5121
 Project: Lucene - Core
  Issue Type: Bug
Reporter: Hoss Man
Priority: Blocker
 Fix For: 4.4

 Attachments: disk_docvals_bug.tgz


 On the #solr irc channel, user adityab reported a problem trying to upgrade 
 an existing solr instance from 4.2.1 to the 4.4 RC0 code.  The specific error 
 he reported was a CorruptIndexException: invalid type: 65 from 
 DiskDocValuesProducer.readFields using an MMapIndexInput.  While waiting for 
 more details from adityab on reproducible configs, i attempted to do a 
 trivial test using simple solr configs based on the docvals test configs, and 
 was able to trigger a slightly different error in which 
 DiskDocValuesProducer.readFields caused an EOFException.
 The problem reproduces on both solr init, and when