Archives of the TeradataForum
Message Posted: Fri, 28 Apr 2006 @ 19:01:30 GMT
| Subj: | | Re: TPUMP Blocking on multiple sessions |
| |
| From: | | Geoffrey Rommel |
| | In the worst case, the blocking can be so severe as to deadlock. Options to deal with it include ... 6) UDFs wrapped in views to transform
the bad PI values into a value/datatype that hashes better. | |
Sorry for the delay, but I'm just now digesting this thread.
Bob, could you elaborate on this 6th option? It seems you are suggesting something like this:
create table T1 ... (bad primary index);
create view V1 as
select udf(PI fields) as better_pi_value,
...other fields ... from T1
with check option;
and then MultiLoading to V1 rather than T1.
Is this a correct interpretation? And how does this help with the blocking in TPump? Sooner or later, the new row has to get to the proper AMP,
and then the blocking is still going to happen ... or is it?
| 
