You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@falcon.apache.org by "Rishav Rohit (JIRA)" <ji...@apache.org> on 2015/09/29 16:48:04 UTC
[jira] [Updated] (FALCON-1498) Add more parameters for HCatalogue
input feed
[ https://issues.apache.org/jira/browse/FALCON-1498?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Rishav Rohit updated FALCON-1498:
---------------------------------
Description:
Add more parameters like falcon_<input_name>_partitions for HCatalogue input feed.
A use case which I can think of now is described below -
Suppose there are two date partitioned input table and you want to join these tables on date partition column -
{quote}
INSERT OVERWRITE TABLE ${falcon_output_database}.${falcon_output_table} PARTITION (${falcon_output_partitions_hive})
SELECT "join_count", count(*)
FROM ${falcon_input1_database}.${falcon_input1_table} a,
${falcon_input2_database}.${falcon_input2_table} b
-- these parameters are not available
WHERE a.${falcon_input1_partitions}
AND b.${falcon_input2_partitions};
--- workaround, use output feed parameters
WHERE a.${falcon_output_partitions}
AND b.${falcon_output_partitions}
{quote}
Some other workaround is also possible, but presence of more parameters would help.
was:
Add more parameters like falcon_<input_name>_partitions for HCatalogue input feed.
A use case which I can think of now is described below -
Suppose there are two date partitioned input table and you want to join these tables on date partition column -
INSERT OVERWRITE TABLE ${falcon_output_database}.${falcon_output_table} PARTITION (${falcon_output_partitions_hive})
SELECT "join_count", count(*)
FROM ${falcon_input1_database}.${falcon_input1_table} a,
${falcon_input2_database}.${falcon_input2_table} b
-- these parameters are not available
WHERE a.${falcon_input1_partitions}
AND b.${falcon_input2_partitions};
--- workaround, use output feed parameters
WHERE a.${falcon_output_partitions}
AND b.${falcon_output_partitions}
Some other workaround is also possible, but presence of more parameters would help.
> Add more parameters for HCatalogue input feed
> ---------------------------------------------
>
> Key: FALCON-1498
> URL: https://issues.apache.org/jira/browse/FALCON-1498
> Project: Falcon
> Issue Type: Improvement
> Components: feed
> Affects Versions: 0.6
> Environment: HDP-2.2.0
> Reporter: Rishav Rohit
> Priority: Minor
>
> Add more parameters like falcon_<input_name>_partitions for HCatalogue input feed.
> A use case which I can think of now is described below -
> Suppose there are two date partitioned input table and you want to join these tables on date partition column -
> {quote}
> INSERT OVERWRITE TABLE ${falcon_output_database}.${falcon_output_table} PARTITION (${falcon_output_partitions_hive})
> SELECT "join_count", count(*)
> FROM ${falcon_input1_database}.${falcon_input1_table} a,
> ${falcon_input2_database}.${falcon_input2_table} b
> -- these parameters are not available
> WHERE a.${falcon_input1_partitions}
> AND b.${falcon_input2_partitions};
> --- workaround, use output feed parameters
> WHERE a.${falcon_output_partitions}
> AND b.${falcon_output_partitions}
> {quote}
> Some other workaround is also possible, but presence of more parameters would help.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)