AMPS and Spooling

[Liam1]

Hello,

When I am inserting lots of data into an empty table
I have noticed that under my user ID, that when the actual insert is taking place, our Teradata box seems to remove the data one AMP at a time.
AMP Spool
1 100
2 100
3 100
4 0
5 0
etc

The data seems to pull off the AMPs in this way. Is this correct?
I thought Teradata was a Parrallel Processing System, ours is definately multi-node and hence should it put all AMPs onto disk at the same time, and not stage it as above?

Thanks,

Liam1

 

[dnoeth]

How did you notice that strange behaviour? Did you run a query repeatedly aganst dbc.diskspace several within several seconds?
I've never seen that before...

Dieter

 

[Liam1]

Dieter,

It also does this when picking data up from the disks and putting it into spool.

It is a data (Set) table with a unique PI.

Our system has always done it...

Liam1