Previous Release Notes : R2014x.SP4
 
R2014x.SP4
 
Improvements
Upgrade notes
Major fixed issues
Other fixed issues
The R2014x.SP4 service pack, is a maintenance release that fixes major issues.
For information about new features, see the online help: About this Release > What's New .
Improvements
Platform
The number of old configuration versions stored in a cvdiag can be limited using a configurable time frame (#20364).
Reducing the cvdiag size makes it easier to provide it to the EXALEAD Support team.
Indexing
DirectIO option is available for index slots (IndexConfig / WriteSlotConfig) (#20575).
Dictionary Builder releases native memory explicitly (#22327).
It reduces memory allocation failure risks for large dictionaries that are built often.
Document directives can be used as conditions within the analysis pipeline (#22297)
The number of analysis jobs pending on import is returned by the getBuildGroupStatus method (#22369)
Search
Hit content for multivalued fields can be limited using the Value selection operation. To add this operation in the Administration Console, go to Search logics > Hit Content > Metas, click Add operation and select Value selection (#22058). For more information, see the Technical Reference Guide > Organizing Search Results > Defining Hit Content > Configure value selection for metas.
Exclude refine on a facet with disjunctive refinement policy is allowed (#22378).
Dynamic date facets performances have been improved (#21317)
Mashup
The CloudView Feed security parameter is now evaluated (#22417)
A new version of the Highchart library (4.0.3) is provided for the Mashup UI (#22120)
Check
highcharts.com for more information.
Documentation
Business Console User Guide: the "Semantic Resources" section has been entirely reviewed.
Upgrade notes
Supported platforms
R2014x was the last CloudView version to support RedHat 5. R2015x will support RedHat 6 only.
Changes that can break things
Change which can break the compatibility of plugins. In previous CloudView versions, the class loading model was an uncommon mix between the java standard mode (where the parent classloader is checked first) and the webapp "reverse" mode.
In 2014x.SP2, we have replaced this mode by two distinct modes:
o The default mode is the java standard mode, because it is the closest to the previous mode.
However, as it is slightly different, you may encounter issues where classes for external libraries embedded in your plugins are not used as they were before (and the versions shipped in CloudView are used instead).
o The reverse mode should help to fix this issue. To activate this mode in your plugin:
1. Edit the cvplugin.properties file
2. Set the plugin.reverse-classloading property to true (plugin.reverse-classloading=true)
Note: You can also edit the
@CVPlugin class with the following annotation:
@CVPluginClassLoading(reverseClassLoading=true)
Major fixed issues
Ticket
Title
Existing since
Component
#22395
Fix correct detection of IE11 on Mashup-UI
Historical
Mashup-UI & API
#22227
MEL cache is not limited in size, Query Builder can cause memory leaks on searchserver
V6R2014x.SP2
Mashup-UI & API
#22406
Aggregation Builder does not output aggregation when fails or empty value
V6R2014x.SP3
Mashup-UI & API
#22355
Some tar file makes File system connector crash
V6R2014x.SP3
Connector File Systems
#22372
Cannot enrich documents pushed by ENOVIA SBA Connector with SDC directives
V6R2014x.SP2
Structured Data Consolidation
#22168
Cannot use addParts aggregation builtin in DSL
V6R2014x.SP3
Structured Data Consolidation
#22174
Object graph cycles may lead to infinite loop
V6R2013
Structured Data Consolidation
#22281
deleteDocumentRootPath does not delete sdc nodes
V6R2014x.SP3
Structured Data Consolidation
#22356
Index warmup does not work in some cases
V6R2014x.SP3
Core
#20251
Crash in containers API: CollectionContainer_fetchElement when indexing rar files
V6R2012x
Core
#22310
Gateway shouldn't crash upon publish dictionary failure
V6R2014x.SP3
Core
#20867
Crash in analysis pipeline with Basis Tech
V6R2014x
Indexing
#20251
Crash of Indexing Server during incremental build on win64 2012
V6R2014 SP2
Indexing
#21773
Crawler documents are not cached in the indexing server
V6R2014 GA
Indexing
#22519
DocumentCache put Document in memory on delete action
V6R2014x.SP1
Indexing
#21678
Unexisting collection db file should not be fatal and crash the gateway when trying to rotate
Historical
Reporting
#21872
Japanese queries may return documents not matching the query (wrong expansion)
Historical
Search
#22105
Use of business console synonyms lead to a crash of slice in proximity processor
Historical
Search
#22234
Dynamic date facet with 1 second missing intervals crash search server
V6R2014x.SP2
Search
#22447
Suggest from query index should not take into account searchlogic limits
V6R2014x.SP1
Search
#22538
Impossible to retrieve values from a Dynamic property
V6R2014x.SP3
Search
#21589
HTTPS Support for PAPI not working
V6R2014x.SP2
Security
#22160
cvconsole & cvadmin not working when MAMI is HTTPS
V6R2014x.SP3
Security
#21753
Mashup Builder not working with a Search API secured with login/password
V6R2014x.SP1
Security
Other fixed issues
Ticket
Title
Existing since
Component
#22472
Display of authentication error messages in MashupUI
V6R2014x.SP3
Mashup-UI & API
#22318
Search server crash on invalid numerical facet label
V6R2014x.SP3
Admin-UI
#22381
Refuse UQL queries not suitable for alerting like #all directly at creation
V6R2014x.SP3
Alerting
#22443
Notes connector does not extract correctly numerical fields
V6R2014x.SP3
Connectors / Lotus Notes
#22136
VC redist 2012 is systematically re-installed
V6R2014x.SP2
Core
#14951
Bad generation of nghostid on aix
V6R2013.SP1
Core
#22067
Taskqueue cannot store files > 2G
V6R2014
Indexing
#18547
incorrect count for disjunctive facets when using collapsing
V6R2013x.SP2
Search
#20998
Long query returns error 500
V6R2014x.SP3
Search
#21727
FUZZYAND expansion is incorrect with a synonym when there are terms following the synonym
V6R2014x.SP1
Search
#22006
Adding meta operations through add_hit_meta does not work
V6R2014.SP1
Search
#22112
Sort/Collapsing time in searchserver not accounted correctly in answer cpu stats
V6R2014x
Search
#22341
Consume less ram in index/searchserver for secured suggest based on categories
V6R2014x
Search
#22617
Duplicated hits received with streaming when a timeout occurs
V6R2014x.SP3
Search
#22503
Boosted Hits OK but wrong score sorting
V6R2014x.SP1
Search