Release Notes for Patch Release 2.7.0.1847 (26-Nov-18)

This is the 2.7.0.1847 Patch Release of Splice Machine, which incorporates updates made since the Patch Release on Nov 19, 2018.

Updates in This Release

Update Type JIRA-ID Description
Bug Internal Major compaction runs in control and never finishes
Bug Internal Flush detection in MemstoreAwareObserver is very brittle
Bug Internal NPE for query on partitioned orc table with predicate where partitioning column is not the last column
Bug Internal Poor write throughput when many Spark tasks writing at the same time
Bug Internal Full distributed standalone splicemachine breaks because not exclude core-site.xml in hbase-sql-*-test.jar
Bug Internal NullPointerException with NestedLoop join and IndexScan.
Bug Internal Make KillOperationIT more robust
Bug Internal TypeDescriptorImpl is not supported on SparkDataset broadcast join implementation
Bug Internal ImportDefaultValueIT fails if SYS."getColumns" gets executed in Spark
Bug SPLICE-2270 Preempted insert operations cause very poor performance overall
Bug SPLICE-2277 TransactionResolver prevents RegionServer from exiting gracefully
Improvement Internal Add Spark Adapter API to split table based on sampling a custom DataFrame
Improvement Internal When inserting data shuffle the partitions to avoid hotspotting
New Feature Internal Add rpms to ease install on MapR 6.0
New Feature SPLICE-2278 NPE for query with WITH clause
Task Internal Q20 failed in TPC-H 100g weekly run
Task Internal Increase number of deploy retries

For a full list of JIRA’s for the Community/Open Source software, see https://splice.atlassian.net