You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-issues@hadoop.apache.org by "Hudson (JIRA)" <ji...@apache.org> on 2016/06/07 23:13:21 UTC
[jira] [Commented] (HADOOP-9956) RPC listener inefficiently assigns
connections to readers
[ https://issues.apache.org/jira/browse/HADOOP-9956?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15319655#comment-15319655 ]
Hudson commented on HADOOP-9956:
--------------------------------
FAILURE: Integrated in HBase-Trunk_matrix #1007 (See [https://builds.apache.org/job/HBase-Trunk_matrix/1007/])
HBASE-15948 Port "HADOOP-9956 RPC listener inefficiently assigns (stack: rev e0b70c00e74aeaac33570508e3732a53daea839e)
* hbase-server/src/test/java/org/apache/hadoop/hbase/ipc/AbstractTestIPC.java
* hbase-server/src/main/java/org/apache/hadoop/hbase/ipc/MetricsHBaseServerWrapperImpl.java
* hbase-server/src/main/java/org/apache/hadoop/hbase/ipc/RpcServer.java
* hbase-hadoop-compat/src/main/java/org/apache/hadoop/hbase/ipc/MetricsHBaseServerSource.java
* hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/SimpleRpcSchedulerFactory.java
> RPC listener inefficiently assigns connections to readers
> ---------------------------------------------------------
>
> Key: HADOOP-9956
> URL: https://issues.apache.org/jira/browse/HADOOP-9956
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: ipc
> Affects Versions: 2.0.0-alpha, 3.0.0-alpha1
> Reporter: Daryn Sharp
> Assignee: Daryn Sharp
> Fix For: 0.23.10, 2.3.0
>
> Attachments: HADOOP-9956.branch-23.patch, HADOOP-9956.patch, HADOOP-9956.patch
>
>
> The socket listener and readers use a complex synchronization to update the reader's NIO {{Selector}}. Updating active selectors is not thread-safe so precautions are required.
> However, the current locking choreography results in a serialized distribution of new connections to the parallel socket readers. A slower/busier reader can stall the listener and throttle performance.
> The problem manifests as unexpectedly low cpu utilization by the listener and readers (~20-30%) under heavy load. The call queue is shallow when it should be overflowing.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-issues-help@hadoop.apache.org