 |
 |
Archives of the TeradataForum
Message Posted: Wed, 17 Dec 2003 @ 15:06:42 GMT
| Subj: | | Fastexport conundrum.. |
| |
| From: | | Meade, Cornelius |
All,
I am looking at a fastexport script that is being run on two separate TD systems in our environment that are largely identical and
contain the same data. The export task runs in about 1 hour when running to one box and takes several days running against the other. In
both cases, the fastexport task is being run from the same non-TPA node. All of the boxes in question are on the same network segment so
network weirdness is not suspected at this time. An examination of the output from the respective export task logs shows that the disparity
in the performance of the two tasks seems to occurring in the phase of the export task where the spooled data blocks are being written out
to the data file. For the first system, this process occurs at a rate in excess of 4,000+ blocks per minute. For the second system, this
rate is about 15 blocks per minute!. Since the amount of data to be exported is about 150,000 blocks in both cases, this would definitely
seem to be the culprit. There must be some system setting or parameter involved here but for the life of me I can not think of what that
might be as yet......Any ideas?
Cornelius Meade III
EDS - AVCS CRM
| |
