Greenplum loader doesn't write data to pipe file

1
votes

We use Informatica to load data into greenplum DB through GP loader.In a single mapping we have parallel flows to insert and update the data in same Greenplum target. Insert is taking a lot of time and finally the workflow has to be aborted.

gploader is creating two yaml files, one for insert and second for update. update will be waiting for the insert process to be completed. But the insert process is taking a lot of time even for 100 records and it never ends, we have to kill the process finally.

Sample YAML File:

%YAML 1.1
---
VERSION: 1.0.0.1

DATABASE: abcdgp_dev
USER: abcs
PASSWORD: srvb&34
HOST: 3.565.785.345
PORT: 5432
GPLOAD:
   INPUT:
     - SOURCE:
        FILE:
          - ../infa_shared/Temp/jaroswind_salesrep_mas_1_pipe
     - COLUMNS:
       - "salesrep_key":
       - "salesrep_id":
       - "salesrep_name":
   - FORMAT: CSV
     - DELIMITER: "\x24"
     - ESCAPE: '/'
     - NULL_AS: '/N'
     - QUOTE: '^'
     - ENCODING: utf8
     - ERROR_LIMIT: 500
     - ERROR_TABLE: jaroswind_err.err_SALESREP_MAS
   OUTPUT:
     - TABLE: jaroswind.salesrep_mas
     - MODE: INSERT
   PRELOAD:
     - REUSE_TABLES: True

please suggest a solution.

1
informaticagreenplum

1 Answers

0
votes

Update is locking the table in between insert