Archives of the TeradataForum
Message Posted: Mon, 19 Jun 2006 @ 17:11:01 GMT
We set up Query Log. I submitted a query as follows,
Select a.UseName, a.defaultDatabase, a.starttime, b.collecttimestamp, b.sqltextInfo from dbc.qryLog a join dbc.qryLogSQL b on a.ProcID = b.ProcID and a.QueryID = b.QueryID Where b.CollectTimeStamp < a.starttime Order by a.starttime, a.username;
I found thousands of records. Ex,
StartTime CollectTimeStamp 2006-02-23 05:12:05.80 2006-02-03 00:06:09 Select . from dbcmgr.TMGRSchedule where .
There were also records generated by other users.
How can the start time be earlier than collect time? What are their definitions? Are there any bugs?
We use Teradata V26R. Teradata SQL assistant, Admin, and Manger are all 7.1. From Teradata Admin, I used Tools/Query Logging ., it did not display any information. All the fields are empty except (1) SQL Length 200, (2) Threshold 5. I did not set up the query logging. How to check the set up for query logging? I do not know how long these records will be kept in the query log tables.
What is the default time period for these records to stay in the tables? Are these tables purged by different processes? How are ProcID and QueryID reused?
Have anyone seen SQLs logged under wrong usernames in the result sets from this kind of queries? I know there are many experts in this forum who are willing to help and share knowledge.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|