[ https://issues.apache.org/jira/browse/HBASE-10531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13937765#comment-13937765 ]
ramkrishna.s.vasudevan commented on HBASE-10531: ------------------------------------------------ Did some benchmarks Mainly took a reading with single node and ran YCSB for pure reads (gets) and sequential scans. The gets were averaged over 50 runs Without patch Throughput : 1418.42019 ops/sec With patch Throughput : 1422.117704 For scans averaged over 15 runs Without patch Throughput :351.0178049 With patch Throughput : 351.4325864 Every run tries to perform 50000 operations. So overall I don't find much difference in the working of reads with and without patch. I have take the plain case(no encoding). I can add results with some encodings like Fast_diff also. > Revisit how the key byte[] is passed to HFileScanner.seekTo and reseekTo > ------------------------------------------------------------------------ > > Key: HBASE-10531 > URL: https://issues.apache.org/jira/browse/HBASE-10531 > Project: HBase > Issue Type: Sub-task > Reporter: ramkrishna.s.vasudevan > Assignee: ramkrishna.s.vasudevan > Fix For: 0.99.0 > > Attachments: HBASE-10531.patch, HBASE-10531_1.patch, > HBASE-10531_2.patch, HBASE-10531_3.patch, HBASE-10531_4.patch, > HBASE-10531_5.patch > > > Currently the byte[] key passed to HFileScanner.seekTo and > HFileScanner.reseekTo, is a combination of row, cf, qual, type and ts. And > the caller forms this by using kv.getBuffer, which is actually deprecated. > So see how this can be achieved considering kv.getBuffer is removed. -- This message was sent by Atlassian JIRA (v6.2#6252)