hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erik Holstad (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-1249) Rearchitecting of server, client, API, key format, etc for 0.20
Date Tue, 28 Apr 2009 05:15:31 GMT

    [ https://issues.apache.org/jira/browse/HBASE-1249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12703498#action_12703498

Erik Holstad commented on HBASE-1249:

Try to answer the questions:
DeleteRow with square brackets only mean that it is a call from the client side but is broken
up into multiple deleteFamilies on the server side, same thing as the type deleteRow.
The sort order is the same as for the Kv.Comparator in KeyValue, if deletes are sorted above
puts, have to check.
Yes, that is exactly what that means. Deletes are taken care of directly in memcache by calling
tailMap with the deleteKeyValue and remove all the affected puts and deletes, so "smaller"
deletes are removed in case of a "bigger" one that comes in afterwards.
In case of a merge deletes needs to be taken care of, so that we don't loose that rule that
deletes only apply to the files after. So for a minor compaction we save all the deletes unless
a "bigger" on comes in first and remove all the affected puts for those deletes.
The DeleteFamily type is moved down through all the file just like the other deletes, but
stored in a separate place and needs to be checked for every delete to see if it applies to
the current KeyValue.

> 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
>         Attachments: HBASE-1249-Example-v1.pdf, HBASE-1249-Example-v2.pdf, HBASE-1249-GetQuery-v1.pdf,
HBASE-1249-GetQuery-v2.pdf, HBASE-1249-GetQuery-v3.pdf, HBASE-1249-GetQuery-v4.pdf, HBASE-1249-StoreFile-v1.pdf,
> 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,

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message