[ https://issues.apache.org/jira/browse/HBASE-13109?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14344810#comment-14344810 ]
ramkrishna.s.vasudevan commented on HBASE-13109: ------------------------------------------------ The optimize() logic makes sense. I think particularly it is going to be useful when there is one version of a cell and the filter/trackers say SEEK_TO_ROW/COL. Changing the HFileBlock Index to Cell is fine unless you have a concern on the number of objects being created and thrown away. In that case we may have to have a different approach but for the case of NO_INDEX_KEY-we cannot go with '==' check. In this case that is not there. > Make better SEEK vs SKIP decisions during scanning > -------------------------------------------------- > > Key: HBASE-13109 > URL: https://issues.apache.org/jira/browse/HBASE-13109 > Project: HBase > Issue Type: Bug > Reporter: Lars Hofhansl > Assignee: Lars Hofhansl > Priority: Minor > Attachments: 13109-0.98-v4.txt, 13109-trunk-v2.txt, > 13109-trunk-v3.txt, 13109-trunk-v4.txt, 13109-trunk-v5.txt, 13109-trunk.txt, > nextIndexKVChange_new.patch > > > I'm re-purposing this issue to add a heuristic as to when to SEEK and when to > SKIP Cells. This has come up in various issues, and I think I have a way to > finally fix this now. HBASE-9778, HBASE-12311, and friends are related. > --- Old description --- > This is a continuation of HBASE-9778. > We've seen a scenario of a very slow scan over a region using a timerange > that happens to fall after the ts of any Cell in the region. > Turns out we spend a lot of time seeking. > Tested with a 5 column table, and the scan is 5x faster when the timerange > falls before all Cells' ts. > We can use the lookahead hint introduced in HBASE-9778 to do opportunistic > SKIPing before we actually seek. -- This message was sent by Atlassian JIRA (v6.3.4#6332)