Skip to content

Cask Data Application Platform - 4.3.1
Compare
Choose a tag to compare
@prinam prinam released this 25 Sep 19:47
· 8852 commits to develop since this release
v4.3.1
c3a1c6f

New Features

  • Adds new visualization tool to give insights about data prepped up in data preparation tool. (CDAP-12592)

  • Adds a way to trigger invalid transaction pruning via a REST endpoint. (CDAP-12620)

  • Adds UI to make HTTP request in CDAP. (CDAP-12595)

Improvements

  • Added a downgrade command to the pipeline upgrade tool, allowing users to downgrade pipelines to a previous version. (CDAP-12598)

  • Improved memory usage of data pipeline with joiner in mapreduce execution engine. (CDAP-12541)

  • Added ability to select/clear all the checkboxes for Provided runtime arguments (CDAP-12176)

  • Fixed a performance issue with the run record corrector (CDAP-12646)

  • Added a capability to configure program containers memory settings through runtime arguments and preferences (CDAP-12380)

  • Applies the extra jvm options configuration to all task containers in MapReduce (CDAP-8499)

  • Fixed a classloader leakage issue when PySpark is used in local sandbox (CDAP-12546)

  • Ability to list the datasets based on the set of dataset properties (CDAP-12593)

Bug Fixes

  • MapReduce Task-related metrics will be emitted from individual tasks instead of MapReduce driver. (CDAP-12645)

  • Fixed the filter if missing flow in the UI to also apply on null values in addition to empty values. (CDAP-12628)

  • Fixed the fill-null-or-empty directive to allow spaces in the default value (CDAP-12612)

  • Fixed a bug that authorization cannot be turned on if kerberos is disabled (CDAP-12588)

  • Fixed an issue that caused the pipeline upgrade tool to upgrade pipelines in a way that would cause UI failures when the upgraded pipeline is viewed. (CDAP-12578)

  • Spark compat directories in the system artifact directory will now be automatically checked, regardless of whether they are explicitly set in app.artifacts.dir. (CDAP-12577)

  • Added option to enable/disable emitting program metrics and option to include or skip task level information in metrics context. This option can be used with scoping at program and program-type level similar to setting system resources with scoping. (CDAP-12570)

  • Improved error messaging when there is an error while in publishing metrics in MetricsCollection service. (CDAP-12569)

  • Fixed a bug that CDAP is not able to clean up and aggregate on streams in an authorization enabled environment (CDAP-12567)

  • Fixed log message format to include class name and line number when logged in master log (CDAP-12559)

  • Adds various improvements to the transaction system, including the ability to limit the size of a transaction's change set; better insights into the cause of transaction conflicts; improved concurrency when writing to the transaction log; better handling of border conditions during invalid transaction pruning; and ease of use for the transaction pruning diagnostic tool. (CDAP-12526)

  • Fix the units for YARN memory stats on Administration UI page. (CDAP-12495)

  • Fixed a bug where the app detail contains entity information that the user does not have any privilege on (CDAP-12482)

  • Fixed preview results for pipelines with condition stages (CDAP-12476)

  • Fixed a bug that caused failures for Hive queries using MR execution engine in CM 5.12 clusters. (CDAP-12457)

  • Fixes an issue where transaction coprocessors could sometimes not access their configuration. (CDAP-12454)

  • UI: Add ability to view payload configuration of pipeline triggers (CDAP-12451)

  • Fixed a bug that the cache timeout was not changed with the value of security.authorization.cache.ttl.secs (CDAP-12441)

  • Fixed an issue with not able to use HiveContext in Spark (CDAP-12415)

  • Added the authorization policy for adding/deleting schedules (CDAP-12387)

  • Fixes an issue where the transaction service could hang during shutdown. (CDAP-12377)

  • Fixed issue where loaded data was not consistently rendering when navigating to Data Preparation from other parts of CDAP. (CDAP-12333)

  • Improves the performance of HBase operations when there are many invalid transactions. (CDAP-12314)

  • Improves a previously misleading log message. (CDAP-12240)

  • Fixed an issue that hive query may failed if the configuration has too many variables substitution. (CDAP-7651)

  • Added mechanism to clean up local dataset if the workflows creating them are killed (CDAP-7243)

  • Improved the error message in the case that a kerberos principal is deleted or keytab is invalid, during impersonation. (CDAP-7049)

Assets 2
Loading