[
https://issues.apache.org/jira/browse/HBASE-1249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12682758#action_12682758
]
Erik Holstad commented on HBASE-1249:
-------------------------------------
Some more thoughts:
DeleteRow should not exist on the server side but broken up into
DeleteFamily for all the families for that region.
Client side
DeleteRow
DeleteFamily
DeleteColumn
Delete
Ts for the deletes should be "now" for DeleteRow and DeleteFamily, an optional
on for DeleteColumn that will delete everything after that ts,
if not specified it will use "now" and Delete should always take a ts for the
put in question and will only react to exact matches of that ts.
and at the server we have
DeleteFamily
DeleteColumn
Delete
I like the idea of having a special ts for the bigger Deletes so you would know
what to get from memCache and where to seekTo in the store,
but maybe an even better option is to have a separate delete set that can be
read in if needed to go to store file for more data.
So for every delete that comes in, it first deletes everything in memCache that
is related to it and then it is added to the deleteSet.
this means that everything that is in memCache is valid data, and you only need
to take deletes into consideration if you are going to the stores.
> Rearchitecting of server, client, API, key format, etc for 0.20
> ---------------------------------------------------------------
>
> Key: HBASE-1249
> URL: https://issues.apache.org/jira/browse/HBASE-1249
> Project: Hadoop HBase
> Issue Type: Improvement
> Reporter: Jonathan Gray
> Priority: Blocker
> Fix For: 0.20.0
>
>
> To discuss all the new and potential issues coming out of the change in key
> format (HBASE-1234): zero-copy reads, client binary protocol, update of API
> (HBASE-880), server optimizations, etc...
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
