 |
 |
Archives of the TeradataForum
Message Posted: Mon, 11 Aug 2003 @ 20:32:48 GMT
| Subj: | | Re: Fastexport initial phase: no disk acess? |
| |
| From: | | Adriano Rodrigues Ferreira |
| | FastExport will not do any disk I/O until after it competes your SELECT. It will not do your SELECT until it has established the
SESSIONS via logon processing. | |
What does it mean? To complete my SELECT, the result set is mounted on memory or just a kind of procedure that will generate the result
set (for example, to be used when fastexport starts to write on disk)? The machine I am working on has 6 AMPs and the script establish the
number of SESSIONS as 30. However, in log files, only 6 sessions are established. Why?
| | If you are using a lot of SESSIONS that can take a while to complete. | |
I will take a look at the manual of FastExport and try to learn something about how to choose this number. I am actually using scripts
written by someone else and I thought they would work seamlessly. As this is not the case, I am working to get a better comprehension of the
tradeoffs involved.
| | Hopefully this gives you something to think about regarding your script and a better understanding of the required
processing. | |
For example, why 486 awtmain processes and for how long I have been watching them, only six were waken up and did something? Teradata is
a deep mistery to me. I thought that the system here is a little twisted and would need some tweaking before working satisfactorily.
Regards,
Adriano.
| |
