You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by "Clint Morgan (JIRA)" <ji...@apache.org> on 2008/10/27 21:21:44 UTC
[jira] Created: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
config ipc.server.tcpnodelay is no loger being respected
--------------------------------------------------------
Key: HADOOP-4525
URL: https://issues.apache.org/jira/browse/HADOOP-4525
Project: Hadoop Core
Issue Type: Bug
Components: ipc
Affects Versions: 0.18.1
Reporter: Clint Morgan
I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (HADOOP-4525) config
ipc.server.tcpnodelay is no loger being respected
Posted by "Chris Douglas (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643092#action_12643092 ]
chris.douglas edited comment on HADOOP-4525 at 10/27/08 4:10 PM:
-----------------------------------------------------------------
This was my fault. HADOOP-2645 knocked the patch for HADOOP-2232 out of sync and I must not have noticed when applying it. Apologies.
I don't think this needs a unit test.
was (Author: chris.douglas):
This was my fault. HADOOP-2643 knocked the patch for HADOOP-2232 out of sync and I must not have noticed when applying it. Apologies.
I don't think this needs a unit test.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Clint Morgan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Clint Morgan updated HADOOP-4525:
---------------------------------
Status: Patch Available (was: Open)
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643207#action_12643207 ]
Hudson commented on HADOOP-4525:
--------------------------------
Integrated in Hadoop-trunk #645 (See [http://hudson.zones.apache.org/hudson/job/Hadoop-trunk/645/])
. Fix ipc.server.ipcnodelay originally missed in in HADOOP-2232.
Contributed by Clint Morgan.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Assignee: Clint Morgan
> Fix For: 0.19.0
>
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Clint Morgan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Clint Morgan updated HADOOP-4525:
---------------------------------
Attachment: hadoop-4525.patch
This fixes it
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Raghu Angadi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643046#action_12643046 ]
Raghu Angadi commented on HADOOP-4525:
--------------------------------------
Can we have a unit test for this? Since this is a performance problem, it won't be detected next time something like this happens.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Raghu Angadi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643050#action_12643050 ]
Raghu Angadi commented on HADOOP-4525:
--------------------------------------
Also it does not look like the this config is was ever read by the server.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Chris Douglas (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Chris Douglas updated HADOOP-4525:
----------------------------------
Resolution: Fixed
Fix Version/s: 0.19.1
Assignee: Clint Morgan
Hadoop Flags: [Reviewed]
Status: Resolved (was: Patch Available)
I just committed this. Thanks, Clint
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Assignee: Clint Morgan
> Fix For: 0.19.1
>
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Clint Morgan (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643057#action_12643057 ]
Clint Morgan commented on HADOOP-4525:
--------------------------------------
I'm not gonna write a unit test just to make sure a config param is being read, but maybe someone else wants to. (Making the field final helps a bit here)
I don't know how you say "it does not look like the this config is was ever read by the server." It was at least being read in the patch that first introduced it: HADOOP-2232
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Chris Douglas (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643092#action_12643092 ]
Chris Douglas commented on HADOOP-4525:
---------------------------------------
This was my fault. HADOOP-2643 knocked the patch for HADOOP-2232 out of sync and I must not have noticed when applying it. Apologies.
I don't think this needs a unit test.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Raghu Angadi (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12643063#action_12643063 ]
Raghu Angadi commented on HADOOP-4525:
--------------------------------------
> I don't know how you say "it does not look like the this config is was ever read by the server." It was at least being read in the patch that first introduced it: HADOOP-2232
For some reason, it was not committed properly : http://svn.apache.org/viewvc/hadoop/core/branches/branch-0.16/src/java/org/apache/hadoop/ipc/Server.java?r1=615093&r2=615098
i.e. committed patch does not read this config.
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HADOOP-4525) config ipc.server.tcpnodelay is no
loger being respected
Posted by "Chris Douglas (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-4525?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Chris Douglas updated HADOOP-4525:
----------------------------------
Fix Version/s: (was: 0.19.1)
0.19.0
> config ipc.server.tcpnodelay is no loger being respected
> --------------------------------------------------------
>
> Key: HADOOP-4525
> URL: https://issues.apache.org/jira/browse/HADOOP-4525
> Project: Hadoop Core
> Issue Type: Bug
> Components: ipc
> Affects Versions: 0.18.1
> Reporter: Clint Morgan
> Assignee: Clint Morgan
> Fix For: 0.19.0
>
> Attachments: hadoop-4525.patch
>
>
> I was troubleshooting some slow IPC from hbase, and though it may be a Naggles algorithm issue. So I turned on tcpNoDelay on the client and server and this had no affect. Turns out that the "ipc.server.tcpnodelay" setting was no longer being read.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.