[jira] Commented: (HBASE-1845) Discussion issue for multi calls, MultiDelete, MultiGet and MultiPut

["stack (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-1845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12757497#action_12757497 ] 

stack commented on HBASE-1845:
------------------------------

I think this patch looks great Holstad.

Here are some initial comments.

Deprecate the old means of batching in favor of your new method?  You can have the old methods call into your new method?

The check a list is sorted probably takes as long as actual sort?

If only one item in list, avoid sort?  Its an optimization we have elsehwere.

Multi should be renamed.  Name it Rows?  MultiResults should be renamed Results?  Says in class comment that Multi is about Gets but its more general than this?

Row has to be a Writable?  Is it OK making it WritableComparable?  I was going to make Row implement Comparable but thought that it'd narrow our ability to mix it in?  If Row is WritableComparable, then should change how Get, Put, etc. implement removing the Comparable and Writable and just use Row?

Can we batch Puts, Deletes, and Gets?  Or does the batch have to be pure -- all of one type?

Whats the threading story?  HTable has a pool of executors used for batching.  So, this would be a continuation of HTable not being thread-safe.  It looks like HCM should be fine if used by many threads?

Testing of failure mid-put with verification that we recover properly would be nice, especially if could be done in a unit test?












> Discussion issue for multi calls, MultiDelete, MultiGet and MultiPut
> --------------------------------------------------------------------
>
>                 Key: HBASE-1845
>                 URL: https://issues.apache.org/jira/browse/HBASE-1845
>             Project: Hadoop HBase
>          Issue Type: New Feature
>            Reporter: Erik Holstad
>             Fix For: 0.21.0
>
>         Attachments: batch.patch, multi-v1.patch
>
>
> I've started to create a general interface for doing these batch/multi calls and would like to get some input and thoughts about how we should handle this and what the protocol should
> look like. 
> First naive patch, coming soon.

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