[ https://issues.apache.org/jira/browse/PHOENIX-3827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16013405#comment-16013405 ]
Rajeshbabu Chintaguntla commented on PHOENIX-3827: -------------------------------------------------- [~jamestaylor] I am working on finding the root cause of OnDuplicateKeyIT test failure but by the time after [~mujtabachohan] raised PHOENIX-3853 looking at the performance issue. I am wondering how the local indexes are slow. As for the patch we collect the local index updates by removing them from index updates if the table is same and add to the list of ongoing mutations. Then what HBase internally does it for every mutation in the batch it will get the corresponding index mutations acquire row lock and merge these cells to data mutations to write atomically(Then the remaining story is same as normal writes). I don't think this is costlier than normal writes. I have things to suspect 1) while loading data observed splits which might have caused some slowness 2) Or else index updates preparation might have been slow Continuing testing to find the root cause of slowness. > Make use of HBASE-15600 to write local index mutations along with data > mutations atomically > ------------------------------------------------------------------------------------------- > > Key: PHOENIX-3827 > URL: https://issues.apache.org/jira/browse/PHOENIX-3827 > Project: Phoenix > Issue Type: Bug > Reporter: Rajeshbabu Chintaguntla > Assignee: Rajeshbabu Chintaguntla > Fix For: 4.11.0 > > Attachments: PHOENIX-3827.patch, PHOENIX-3827_v2.patch, > PHOENIX-3827_v3.patch > > > After HBASE-15600 we can add mutations of the same table from coprocessors so > we can write local index data along with data mutations so it will be atomic. > This we can do in 4.x-HBase-1.3 version. -- This message was sent by Atlassian JIRA (v6.3.15#6346)