[ https://issues.apache.org/jira/browse/KUDU-749?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16854764#comment-16854764 ]
Todd Lipcon commented on KUDU-749: ---------------------------------- Hi [~ZhangYao]. That makes sense to me -- maybe the POC will expose some issue that we aren't thinking about, but it's worth trying. Another thing you could try is to implement "seek before": eg if you are looking for the latest delta on row 10, seek instead to row 11 ts=0, and then scan left by one entry. This should be possible to implement in O(lgn) time. > Improve performance for zipfian update > -------------------------------------- > > Key: KUDU-749 > URL: https://issues.apache.org/jira/browse/KUDU-749 > Project: Kudu > Issue Type: Improvement > Components: perf, tablet > Affects Versions: Private Beta > Reporter: Todd Lipcon > Priority: Major > Attachments: screenshot-1.png, screenshot-2.png > > > A zipfian 50/50 update/read workload on YCSB gets slower and slower until > it's pretty intolerable (random reads taking 100+ms of CPU). It seems like > all the CPU is spent in DMSIterator::PrepareBatch. We're probably doing > something dumb here - let's look for some low hanging fruit to fix this. -- This message was sent by Atlassian JIRA (v7.6.3#76005)