[
https://issues.apache.org/jira/browse/PHOENIX-2154?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14720635#comment-14720635
]
Ravi Kishore Valeti commented on PHOENIX-2154:
----------------------------------------------
Ex: we have a single map task with 1000 records.
If we do context.write() in map() method, no.of input records for reducer=1000
If we do context.write() in cleanup() method ( this will be called once all
records are processed by map()), by writing some dummy key-value, then no. Of
input records to reduce = no.of map tasks=1.
I tested below cases
- context.write(key,value) in map() method
Result: map input records = 1000
map output records = 1000
Reducer input records = 1000
- context.write(key,value) in cleanup() method
Result: map input records = 1000
map output records = 1
Reducer input records = 1
> Failure of one mapper should not affect other mappers in MR index build
> -----------------------------------------------------------------------
>
> Key: PHOENIX-2154
> URL: https://issues.apache.org/jira/browse/PHOENIX-2154
> Project: Phoenix
> Issue Type: Bug
> Reporter: James Taylor
> Assignee: Ravi Kishore Valeti
> Attachments: IndexTool.java, PHOENIX-2154-WIP.patch,
> PHOENIX-2154-_HBase_Frontdoor_API_WIP.patch,
> PHOENIX-2154-_HBase_Frontdoor_API_v1.patch,
> PHOENIX-2154-_HBase_Frontdoor_API_v2.patch
>
>
> Once a mapper in the MR index job succeeds, it should not need to be re-done
> in the event of the failure of one of the other mappers. The initial
> population of an index is based on a snapshot in time, so new rows getting
> *after* the index build has started and/or failed do not impact it.
> Also, there's a 1:1 correspondence between index rows and table rows, so
> there's really no need to dedup. However, the index rows will have a
> different row key than the data table, so I'm not sure how the HFiles are
> split. Will they potentially overlap and is this an issue?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
