Thanks Ted, This issue came up from a user who upgraded from 0.98 based hbase that didn't have the leak to a 1.x+ based hbase which does have the leak. It is a regression and code hygiene issue so it makes sense to fix this and allow old code to continue to work.
Jon. On Sun, Jun 26, 2016 at 5:01 PM, Ted Malaska <[email protected]> wrote: > If u use bulk put in the spark module it will do the same functionality and > not have this problem. > > > > On Sunday, June 26, 2016, Jonathan Hsieh (JIRA) <[email protected]> wrote: > > > Jonathan Hsieh created HBASE-16117: > > -------------------------------------- > > > > Summary: Fix Connection leak in mapred.TableOutputFormat > when > > spark uses it. > > Key: HBASE-16117 > > URL: https://issues.apache.org/jira/browse/HBASE-16117 > > Project: HBase > > Issue Type: Bug > > Components: mapreduce > > Affects Versions: 2.0.0, 1.3.0, 1.2.2, 1.1.6 > > Reporter: Jonathan Hsieh > > Assignee: Jonathan Hsieh > > Fix For: 2.0.0, 1.3.0, 1.2.2, 1.1.6 > > > > > > Spark seems to instantiate multiple instances of output formats within a > > single process. When mapred.TableOutputFormat (not > > mapreduce.TableOutputFormat) is used, this may cause connection leaks > that > > slowly exhaust the cluster's zk connections. > > > > This patch fixes that. > > > > > > > > -- > > This message was sent by Atlassian JIRA > > (v6.3.4#6332) > > > > > -- > Sent from Gmail Mobile > -- // Jonathan Hsieh (shay) // HBase Tech Lead, Software Engineer, Cloudera // [email protected] // @jmhsieh
