Archives of the TeradataForum
Message Posted: Thu, 08 Mar 2001 @ 17:19:15 GMT
| Subj: || || Re: Abort Session from BTEQ |
| From: || || Geoffrey Rommel |
I didn't ponder the details on this, but if you're running out of spool space even on small queries, it almost always means that the data
is piling up on one AMP (vproc). The most frequent culprit seems to be redistributing the data by some column that contains a lot of NULLs;
all the nulls, of course, are distributed to the same AMP (usu. the last one). But it could be some other value (perhaps during the product
join of Product & Offers).
I suggest that you run the query again and have someone else query DBC.DiskSpace (?) as it is running. This view contains a column,
CurrentSpool, that tells how much spool space is being used on each AMP for the user who is running the query. This will tell you which AMP
is filling up, and that should give you a clue where the problem lies. For instance, if the user running the query is HENRY8, run this as
often as possible while they're running:
/* WARNING: UNTESTED SQL */
select Vproc, CurrentSpool
where DatabaseName = 'HENRY8'
order by 2 desc;