[ https://issues.apache.org/jira/browse/HBASE-18010?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16018772#comment-16018772 ]
Anastasia Braginsky commented on HBASE-18010: --------------------------------------------- If we are already talking about... I wonder about the search for the first and last keys in the CellSet {code} public Cell first() { return this.delegatee.get(this.delegatee.firstKey()); } public Cell last() { return this.delegatee.get(this.delegatee.lastKey()); } {code} As our mapping is from Cell to Cell after this.delegatee.firstKey() or this.delegatee.lastKey() we already have the correct Cell in our hands. Why do we do another logN traversal in this.delegatee.get()? Looks like a waste of time...? Further more, in CellFlatMaps this.delegatee.firstKey() and this.delegatee.lastKey() are working in O(1) so adding O(logN) to it sounds not so good... Any explanations? > Connect CellChunkMap to be used for flattening in CompactingMemStore > -------------------------------------------------------------------- > > Key: HBASE-18010 > URL: https://issues.apache.org/jira/browse/HBASE-18010 > Project: HBase > Issue Type: Sub-task > Reporter: Anastasia Braginsky > > The CellChunkMap helps to create a new type of ImmutableSegment, where the > index (CellSet's delegatee) is going to be CellChunkMap. No big cells or > upserted cells are going to be supported here. -- This message was sent by Atlassian JIRA (v6.3.15#6346)