Quantcast
Channel: Teradata Forums - All forums
Viewing all articles
Browse latest Browse all 27759

Moving window function - response (1) by dnoeth

$
0
0
You should change the update to a merge, which should get a much better plan: MERGE INTO tmp AS tgt USING ( SELECT col1,col2,col3, ST_DT, COALESCE(MAX(ST_DT) OVER (PARTITION BY col1 ORDER BY ST_DT ROWS BETWEEN 1 FOLLOWING AND 1 FOLLOWING) , TIMESTAMP '9999-12-12 00:00:00') AS end_date FROM tmp ) AS src ON tgt.COL1 = src.COL1 AND tgt.COL2 = src.COL2 AND tgt.COL3 = src.COL3 AND tgt.ST_DT = src.ST_DT AND tgt.end_dt = TIMESTAMP '9999-12-12 00:00:00' WHEN MATCHED THEN UPDATE SET end_dt = src.end_date Btw, it should be enough to specify the Primary Key column(s) plus the end_date in the ON clause. Dieter

Viewing all articles
Browse latest Browse all 27759

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>