Releases: gchq/stroom
stroom distribution v5.0-beta.35
- Issue #322 : The index cache and other caching beans should now throw exceptions on
get
that were generated during the creation of cached items
stroom distribution v5.0-beta.34
-
Issue #325 : Query history is now cleaned with a separate job. Also query history is only recorded for manual querying, i.e. not when query is automated (on open or auto refresh). Queries are now recorded on a dashboard + query component basis and do not apply across multiple query components in a dashboard.
-
Issue #323 : Fixed an issue where parser elements were not being returned as 'processors' correctly when downstream of a reader.
-
Issue #322 : Index should now provide a more helpful message when an attempt is made to index data and no volumes have been assigned to an index.
stroom distribution v5.0-beta.33
-
Issue #316 : Search history is now only stored on initial query when using automated queries or when a user runs a query manually. Search history is also automatically purged to keep either a specified number of items defined by
stroom.query.history.itemsRetention
(default 100) or for a number of days specified bystroom.query.history.daysRetention
(default 365). -
Issue #317 : Users now need update permission on an index plus 'Manage Index Shards' permission to flush or close index shards. In addition to this a user needs delete permission to delete index shards.
-
Issue #319 : SaveAs now fetches the parent folder correctly so that users can copy items if they have permission to do so.
stroom distribution v5.0-beta.32
-
Issue #311 : Fixed request for
Pipeline
inmeta
XSLT function. Errors are now dealt with correctly so that the XSLT will not fail due to missing meta data. -
Issue #313 : Fixed case of
xmlVersion
property onInvalidXMLCharFilterReader
. -
Issue #314 : Improved description of
tags
property inBadTextXMLFilterReader
.
stroom distribution v5.1-alpha.1
- Issue #202 : Initial release of the new data retention policy functionality.
stroom distribution v5.0-beta.31
-
Issue #307 : Made some changes to avoid potential NPE caused by session serialisation.
-
Issue #306 : Added a stroom
meta
XSLT function. The XSLT function now exposesFeed
,StreamType
,CreatedTime
,EffectiveTime
andPipeline
meta attributes from the currently processing stream in addition to any other meta data that might apply. To access these meta data attributes of the current stream usestroom:meta('StreamType')
etc. Thefeed-attribute
function is now an alias for themeta
function and should be considered to be deprecated. -
Issue #303 : The stream delete job now uses cron in preference to a frequency.
stroom distribution v5.0-beta.30
- Issue #152 : Changed the way indexing is performed so that a single indexer object is now responsible for indexing documents and adding them to the appropriate shard.
stroom distribution v5.0-beta.29
stroom distribution v5.0-beta.28
-
Issue #295 : Made the help URL absolute and not relative.
-
Issue #293 : Attempt to fix mismatch document count error being reported when index shards are opened.
-
Issue #292 : Fixed locking for rolling stream appender.
-
Issue #292 : Rolling stream output is no longer associated with a task, processor or pipeline to avoid future processing tasks from deleting rolling streams by thinking they are superseded.
-
Issue #292 : Data that we expect to be unavailable, e.g. locked and deleted streams, will no longer log exceptions when a user tries to view it and will instead return an appropriate message to the user in place of the data.
stroom distribution v5.0-beta.27
-
Issue #288 : The error condition 'Expected a new writer but got the same one back!!!' should no longer be encountered as the root cause should now be fixed. The original check has been reinstated so that processing will terminate if we do encounter this problem.
-
Issue #295 : Fixed the help property so that it can now be configured.
-
Issue #296 : Removed 'New' and 'Delete' buttons from the global property dialog.
-
Issue #279 : Fixed NPE thrown during proxy aggregation.
-
Issue #294 : Changing stream task status now tries multiple times to attempt to avoid a hibernate LockAcquisitionException.