[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17182340#comment-17182340 ] Xiaoqiao He commented on HDFS-15448: Thanks [~jianghuazhu] for your works here. Failed unit tested

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181811#comment-17181811 ] Hadoop QA commented on HDFS-15448: -- | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem ||

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181763#comment-17181763 ] jianghua zhu commented on HDFS-15448: - [~hemanthboyina] , thank you. > When starting a DataNode,

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181664#comment-17181664 ] Hemanth Boyina commented on HDFS-15448: --- triggered build > When starting a DataNode, call

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17181660#comment-17181660 ] Xiaoqiao He commented on HDFS-15448: Hi [~jianghuazhu], Please check the failed unit tests if it is

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17170123#comment-17170123 ] Hadoop QA commented on HDFS-15448: -- | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem ||

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17169995#comment-17169995 ] jianghua zhu commented on HDFS-15448: - [~elgoiri] , [~weichiu] , do you agree with this patch

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17153419#comment-17153419 ] jianghua zhu commented on HDFS-15448: - [~hexiaoqiao] , thank you very much.   > When starting a

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17153311#comment-17153311 ] Xiaoqiao He commented on HDFS-15448: LGTM, +1 for [^HDFS-15448.003.patch]. > When starting a

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17153254#comment-17153254 ] Hadoop QA commented on HDFS-15448: -- | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem ||

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17153188#comment-17153188 ] jianghua zhu commented on HDFS-15448: - [~elgoiri] , your suggestions are very helpful to me. Here I

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17152972#comment-17152972 ] Íñigo Goiri commented on HDFS-15448: [~hexiaoqiao] already went through the code paths. We may want

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17152556#comment-17152556 ] jianghua zhu commented on HDFS-15448: - [~elgoiri] , [~linyiqun] , can you give me some more

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149897#comment-17149897 ] Xiaoqiao He commented on HDFS-15448: Hi [~linyiqun], As [~jianghuazhu] mentioned above,

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149772#comment-17149772 ] jianghua zhu commented on HDFS-15448: - [~linyiqun] , thank you for your attention. Do you think the

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149454#comment-17149454 ] jianghua zhu commented on HDFS-15448: - [~linyiqun] , thank you very much for reviewing the code.

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149396#comment-17149396 ] Yiqun Lin commented on HDFS-15448: -- Not sure if it's a right behavior to remove startAll() in

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149251#comment-17149251 ] Hadoop QA commented on HDFS-15448: -- | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem ||

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149164#comment-17149164 ] jianghua zhu commented on HDFS-15448: - [~hexiaoqiao] , thank you very much. I have resubmitted a new

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149152#comment-17149152 ] Xiaoqiao He commented on HDFS-15448: Ping [~weichiu],[~elgoiri],[~linyiqun], would you like to have

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149091#comment-17149091 ] Xiaoqiao He commented on HDFS-15448: {quote}whether your suggestion is to increase synchronization

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17149053#comment-17149053 ] jianghua zhu commented on HDFS-15448: - [~hexiaoqiao] , thank you very much for your suggestions. I

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17148812#comment-17148812 ] Hadoop QA commented on HDFS-15448: -- | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem ||

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17148723#comment-17148723 ] Xiaoqiao He commented on HDFS-15448: Thanks [~jianghuazhu], Good catch here. v001 almost LGTM, just

[jira] [Commented] (HDFS-15448) When starting a DataNode, call BlockPoolManager#startAll() twice.

[ https://issues.apache.org/jira/browse/HDFS-15448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17148684#comment-17148684 ] jianghua zhu commented on HDFS-15448: - The source code and unit tests have been verified, without any