Archives of the TeradataForum

Message Posted: Thu, 10 Jul 2003 @ 13:30:27 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Dave,

thanks it helps a bit. I realize I simplified my example a bit, but I was having trouble articulating what I wanted to know.

So we agree that a user session connected via a particular PE, does not necessarily imply what node (IP) that a user connected with.

What I would like to know is - how do I map back a Teradata session to a node (IP) connection? i.e. what IP did a user use to connect? (that's what I meant by Unix connection). How could I know what IP address a user used to connect? (and before you ask why I'd want to, let's say I had network settings wrong on some of the nodes and I was trying to do problem determination)

For a test, I connected to node 10 via telnet and started BTEQ, within PMON or qrysessn, I can see my session and the PE associated with it, (and it all probability, that PE will not be associated with node 10), in this example is connected via node 1.

Based on the fact that the IP connection (node connected) does not impact which PE connection is made through, why do you think one node will be driven harder than the others? I assumed that the only impact would be that I'd see more network traffic out of that node if large answer sets were being returned.

Ruth Fenwick
Database Services

	<Prev	Next>		<<First	<Prev	Next>	Last>>