Just noting down some of the observations which helped me with day to day work involved with Datastage..
- While compiling a Datastage Job, if it gives error on passive stage then check if APT_CONFIG_FILE is present.
- Teradata API Stage uses only ANSI SQL.
- If reading multiple files from DS Sequence file and all of them have header records, then only header record of first file will be dropped.
- If in between closecmd execution, DB gets disconnected then it will rollback all the transactions.
- In Enterprise stage, Replace will not create table with column type as Unicode but wil l create as Latin only.
- In API Stage, to improve performance increase Transaction size, use parallel mode and give hash partitioning . But don't use parallel mode when there are after SQL in that stage or it is used as source stage.
- In API Stage, if User defined SQL is used then we need to add all the columns in using clause as it doesn't recognize column by column name but by sequence order.
- Ex. Using A, B, C Update Table SET Col1 = :B where col2 = :C ; Here if columns from job are in order D, F, G, H, A, B then :B = F
No comments:
Post a Comment