[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14614105#comment-14614105 ] Karthik Kambatla commented on YARN-3860: bq. That is true only if RM restart is

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605508#comment-14605508 ] Hudson commented on YARN-3860: -- SUCCESS: Integrated in Hadoop-Yarn-trunk-Java8 #243 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605517#comment-14605517 ] Hudson commented on YARN-3860: -- SUCCESS: Integrated in Hadoop-Yarn-trunk #973 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605673#comment-14605673 ] Hudson commented on YARN-3860: -- FAILURE: Integrated in Hadoop-Hdfs-trunk #2171 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605704#comment-14605704 ] Hudson commented on YARN-3860: -- FAILURE: Integrated in Hadoop-Hdfs-trunk-Java8 #232 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605851#comment-14605851 ] Hudson commented on YARN-3860: -- SUCCESS: Integrated in Hadoop-Mapreduce-trunk #2189 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605883#comment-14605883 ] Hudson commented on YARN-3860: -- SUCCESS: Integrated in Hadoop-Mapreduce-trunk-Java8 #241 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14606588#comment-14606588 ] Karthik Kambatla commented on YARN-3860: Also, I wonder if this affects the latency

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14606586#comment-14606586 ] Karthik Kambatla commented on YARN-3860: I am not convinced we need this. The only

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605207#comment-14605207 ] Masatake Iwasaki commented on YARN-3860: Thanks, [~zxu] and [~djp]! rmadmin

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14606870#comment-14606870 ] Junping Du commented on YARN-3860: -- bq. The only reason we don't want both RMs to be

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14607191#comment-14607191 ] Masatake Iwasaki commented on YARN-3860: bq. Today, if both RMs become active, one

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14607196#comment-14607196 ] Masatake Iwasaki commented on YARN-3860: bq. I wonder if this affects the latency

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605073#comment-14605073 ] Junping Du commented on YARN-3860: -- Thanks [~iwasakims] for nice catching and the patch!

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605108#comment-14605108 ] zhihai xu commented on YARN-3860: - yes, it makes sense. Although they are equivalent, it is

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14605131#comment-14605131 ] Hudson commented on YARN-3860: -- FAILURE: Integrated in Hadoop-trunk-Commit #8082 (See

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14604506#comment-14604506 ] Hadoop QA commented on YARN-3860: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ ||

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14604531#comment-14604531 ] Hadoop QA commented on YARN-3860: - \\ \\ | (/) *{color:green}+1 overall{color}* | \\ \\ ||

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14604509#comment-14604509 ] Hadoop QA commented on YARN-3860: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ ||

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14604539#comment-14604539 ] zhihai xu commented on YARN-3860: - [~iwasakims], thanks for working on this issue. This

[jira] [Commented] (YARN-3860) rmadmin -transitionToActive should check the state of non-target node

[ https://issues.apache.org/jira/browse/YARN-3860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14604474#comment-14604474 ] Masatake Iwasaki commented on YARN-3860: HAAdmin#isOtherTargetNodeActive does not