[jira] [Updated] (HIVE-10132) LLAP: Tez heartbeats are delayed by ~500+ ms due to Hadoop IPC client
[ https://issues.apache.org/jira/browse/HIVE-10132?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Gopal V updated HIVE-10132: --- Attachment: HIVE-10132.NOT.A.patch This turns off the client-cache, which uses the socket factory object as the key. But is not the right fix. > LLAP: Tez heartbeats are delayed by ~500+ ms due to Hadoop IPC client > - > > Key: HIVE-10132 > URL: https://issues.apache.org/jira/browse/HIVE-10132 > Project: Hive > Issue Type: Sub-task >Reporter: Gopal V >Assignee: Siddharth Seth > Attachments: HIVE-10132.NOT.A.patch > > > HADOOP-11772 has a clearer bug report of the core issue inside hadoop-common. > Due to the delayed heartbeats reaching the AM, the reducers are losing up-to > a couple of seconds for a 60ms (x10 parallel) mapper + 300ms reducer instead > of finishing the query in under a second. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HIVE-10132) LLAP: Tez heartbeats are delayed by ~500+ ms due to Hadoop IPC client
[ https://issues.apache.org/jira/browse/HIVE-10132?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Gopal V updated HIVE-10132: --- Assignee: Siddharth Seth > LLAP: Tez heartbeats are delayed by ~500+ ms due to Hadoop IPC client > - > > Key: HIVE-10132 > URL: https://issues.apache.org/jira/browse/HIVE-10132 > Project: Hive > Issue Type: Sub-task >Reporter: Gopal V >Assignee: Siddharth Seth > > HADOOP-11772 has a clearer bug report of the core issue inside hadoop-common. > Due to the delayed heartbeats reaching the AM, the reducers are losing up-to > a couple of seconds for a 60ms (x10 parallel) mapper + 300ms reducer instead > of finishing the query in under a second. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
