Monday, May 10, 2010

A SQLNet thing I probably forgot about...

My most recent "ah-hah", or "oh yeah" moment came reading Jonathan Lewis's blog. A very neat "SQL Net compression" detail.

It is one of those things that I cannot remember if it was something I knew but forgot - or just never knew.

His example makes the point nicely, you have to appreciate those little test scripts for that. You can clearly see what a difference an order by might make on a the size of a result set set across the network.

Yet another reason to look at bulk processing - the more reasons the better...


Blogger Chris said....

Same thing here. I might have picked it up on the usenet years ago, there was a vague deja-whatever moment.

On a different note, when registering for the AOUG conference, for a very short time I was wondering if the beard really was off... *g*

Tue May 11, 01:08:00 PM EDT  

Anonymous Alberto Dell'Era said....


do you happen to know whether/where this feature is documented ?

I'm trying to check whether the same is available for client-to-server communications (e.g. batch-inserts); I have been unable to check it so far.


Thu May 13, 05:02:00 AM EDT  

Anonymous Vidar E said....

I've read about and tested this before. Works fine. Can't remember where I read about it though, but I have an idea about it emerging in 10g.

Anyway - this feature also comes nicely into play when doing master-detail-joins: Don't be afraid of repeating values taking up space on the network. (Might be a problem in then receiving applications buffer though - depending on the implementation)

Thu May 20, 03:46:00 AM EDT  


<< Home