[ https://issues.apache.org/jira/browse/HDFS-918?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jay Booth updated HDFS-918: --------------------------- Attachment: hdfs-918-pool.patch Here's the patch with just the filechannel pool improvements. This one showed some decent performance improvements while incorporating existing logic regarding the bits and bytes of block reading -- memoized block and all of that. The multiplexed implementation is currently missing memoized block. Would people be interested in another JIRA focusing on just the pooling improvements (.5ms per read or so)? I could forward-port that to trunk and then continue work on the multiplexing implementation which did seem to provide an additional performance boost. > Use single Selector and small thread pool to replace many instances of > BlockSender for reads > -------------------------------------------------------------------------------------------- > > Key: HDFS-918 > URL: https://issues.apache.org/jira/browse/HDFS-918 > Project: Hadoop HDFS > Issue Type: Improvement > Components: data-node > Reporter: Jay Booth > Fix For: 0.22.0 > > Attachments: hbase-hdfs-benchmarks.ods, hdfs-918-20100201.patch, > hdfs-918-20100203.patch, hdfs-918-20100211.patch, hdfs-918-20100228.patch, > hdfs-918-20100309.patch, hdfs-918-branch20-append.patch, > hdfs-918-branch20.2.patch, hdfs-918-pool.patch, hdfs-918-TRUNK.patch, > hdfs-multiplex.patch > > > Currently, on read requests, the DataXCeiver server allocates a new thread > per request, which must allocate its own buffers and leads to > higher-than-optimal CPU and memory usage by the sending threads. If we had a > single selector and a small threadpool to multiplex request packets, we could > theoretically achieve higher performance while taking up fewer resources and > leaving more CPU on datanodes available for mapred, hbase or whatever. This > can be done without changing any wire protocols. -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.