Qubole Release Notes - 16-Dec-2015

See What is New and List of Changes and Bug Fixes in this Release for this version of QDS release that are mentioned below.

What is New

Frontend Performance Enhancements

These are the frontend performance enhancements in this release:

  • Spriting of images
  • Combining and minifying of JS and CSS files

Higher Concurrency for Spark SQL Commands in Notebooks

Notebooks now support higher concurrency for running Spark SQL commands. Earlier, the concurrency was controlled by setting zeppelin.spark.concurrentSQL to true due to which the maximum concurrency that was achievable was about 7-8. Now, you can set an additional parameter, zeppelin.spark.sql.maxConcurrency to a positive integer. Interpreters ensure that many Spark SQL commands can run in parallel.  

List of Changes and Bug Fixes in this Release

HADOOP-1

Change

HAD-527: Task Logs for retired tasks that are greater than 100 MB are no longer downloadable directly via Qubole UI. They can be retrieved from S3 directly.

 


HADOOP-2

 

Change

Monitor support for the following daemons: ResourceManager, NameNode, Yarn ApplicationTimelineServer, and MapReduce JobHistoryServer, in addition to NodeManager and DataNode.

Fix

Do not timeout after 5 successive retries of a NodeManager failure. This affects node recommissioning.

 

PRESTO

Fix

PRES-518: Re-introducing target location cleanup before CTAS. This caused inconvenience for customers running scheduled jobs as CTAS as once CTAS fails, retries fail with target directory already exists  exception.

Fix

PRES-516: Fix drop table failures. This fixes the issue where drop table command drops the table but then fails with Table does not exist exception.



QDS

 

Fix

UI-2849 - EBS volume count should be 1 for ebsonly instance(c4,m4) only

Fix

ZEP-59: Fix REST API to not update cluster id when name is updated

Fix

QBOL-4831: Resolve 500 error in users signin

Fix

QBOL-4830: Use roles for adhoc script API endpoint

Fix

QBOL-4795: Resolve 500 error in switching account

Fix

UI-2644: Results table do not display column names for the first time

Fix

UI-2688: Added condition to check for stable spot bid percentage only when stable spot purchasing option is checked

Fix

SPAR-806: System user will be able to perform CRUD operations

Fix

UI-2593: Analyze - improve query history list ajax calls

Fix

Query hists table - drop qbol_user_id index. Add index on qbol_user_id and submit_time

Fix

Fix HBase Ganglia to send Metrics to Master DNS

Fix

UI-2892: Cluster selection not required for Data Import and Data Export. Hive Queries allowed to run on Spark cluster.

Change

ZEP-56: Add Hacker new dataset related Example note

 

ZEPPELIN/NOTEBOOK

Fix

ZEP-17: Fix create->cancel->create flow in interpreter page

Fix

ZEP-48: Fix status of paragraphs



Known Issues

 

None



List of Hotfixes Since 8th December 2015


None

Have more questions? Submit a request

Comments

Powered by Zendesk