hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ratandeep Ratti (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HIVE-18410) [Performance][Avro] Reading flat Avro tables is very expensive in Hive
Date Fri, 12 Jan 2018 18:23:00 GMT

     [ https://issues.apache.org/jira/browse/HIVE-18410?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Ratandeep Ratti updated HIVE-18410:
-----------------------------------
    Attachment: HIVE-18410_1.patch

> [Performance][Avro] Reading flat Avro tables is very expensive in Hive
> ----------------------------------------------------------------------
>
>                 Key: HIVE-18410
>                 URL: https://issues.apache.org/jira/browse/HIVE-18410
>             Project: Hive
>          Issue Type: Improvement
>    Affects Versions: 2.3.2
>            Reporter: Ratandeep Ratti
>            Assignee: Ratandeep Ratti
>             Fix For: 2.3.2
>
>         Attachments: HIVE-18410.patch, HIVE-18410_1.patch, profiling_with_patch.nps,
profiling_with_patch.png, profiling_without_patch.nps, profiling_without_patch.png
>
>
> There's a performance penalty when reading flat [no nested fields] Avro tables. When
reading the same flat dataset in Pig, it takes half the time.  On profiling, a lot of time
is spent in {{AvroDeserializer.deserializeSingleItemNullableUnion()}}. The bulk of the time
is spent in GenericData.get().resolveUnion(), which calls GenericData.getSchemaName(Object
datum), which does a lot of instanceof checks.  This could be simplified with performance
benefits. A approach is described in this patch which almost halves the runtime.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message