Archives of the TeradataForum
Message Posted: Wed, 16 Nov 2005 @ 14:42:37 GMT
Subj: | | Re: Excessive Delays between DBQLogTbl's StartTime and FirstStepTime |
|
From: | | Wardius, Mark |
Carrie,
Since we are on the subject of DBQL could you shed any light on why the dbc.dbqlogtbl.CollectTimeStamp values are dated several hours prior to
the dbc.dbqlogtbl.Startime and other time fields? We just did a 10 node upgrade (the day before) and we wanted to see if any activity occurred
during the period the users were revoked. To our amazement when we queried dbc.dbqlogtbl.CollectTimeStamp we found several hundred Microstrategy
type queries. Then to verify our results we filtered on the dbc.dbqlogtbl.starttime field and we saw these results (these results are a subset of
the hundreds of rows returned)
CollectTimeStamp LogonDateTime StartTime FirstStepTime FirstRespTime
11/13/05 7:09 AM 11/13/05 12:32 PM 11/13/05 12:32 PM 11/13/05 12:32 PM 11/13/05 12:32 PM
11/13/05 7:10 AM 11/13/05 12:50 PM 11/13/05 12:50 PM 11/13/05 12:50 PM 11/13/05 12:51 PM
11/13/05 7:10 AM 11/13/05 1:24 PM 11/13/05 1:24 PM 11/13/05 1:24 PM 11/13/05 1:24 PM
11/13/05 7:11 AM 11/13/05 12:54 PM 11/13/05 12:54 PM 11/13/05 12:54 PM 11/13/05 12:57 PM
I know that Teradata is an excellent RDBMS; yet I did not know it was physic. DBQL is now correctly reporting the CollectTimeStamp at
this time, yet in a weird time continuum event our system could predict the future.
Any ideas of what may have happened just after the ten node upgrade??? &-)
Mark Wardius
|