[ https://issues.apache.org/jira/browse/HBASE-15948?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15313402#comment-15313402 ]
stack commented on HBASE-15948: ------------------------------- 002 is after review of current state of hadoop Server... looks like it changed again since HADOOP-9956 using a LinkedBlockingQueue instead of ConcurrentLinkedQueue. Adds new "ipc.server.read.connection-queue.size" config for how many items to queue per Reader instance (this is distinct from "hbase.ipc.server.listen.queue.size" which is how many calls to queue on the socket) > Port "HADOOP-9956 RPC listener inefficiently assigns connections to readers" > ----------------------------------------------------------------------------- > > Key: HBASE-15948 > URL: https://issues.apache.org/jira/browse/HBASE-15948 > Project: HBase > Issue Type: Sub-task > Components: IPC/RPC > Reporter: stack > Assignee: stack > Attachments: HBASE-15948.master.001.patch, > HBASE-15948.master.002.patch > > > Esteban noticed we were missing this upstream issue. Seems to make no > difference in profiling but here is the patch anyways. -- This message was sent by Atlassian JIRA (v6.3.4#6332)