| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> Re: IMP-00020 info
I've been fighting this very error all week -- when my export/import had been working just a week ago. I submitted a TAR, and all the Oracle techie could say was that the dmp file was probably corrupt. I responded warily, saying that five different files would have to have been corrupt -- and why were the other 140 tables importing just fine???
Anyway, I worked too many hours this week to ruin my weekend over this. But first thing Monday morning I'm going to test your workaround (however illogical it seems!!) If it gets me clear, I'll buy you a virtual beer (or another beverage of your choosing!!)
Thanks for sharing the experience. I was feeling like I was rowing a lone boat in Oracle's Twilight Zone.
Joe Cooper
Oracle DBA
Company To Be Determined
(we're being sold, and are supposed to
meet the new owners Monday!!)
Ron Rogers <RROGERS_at_galottery.org> wrote:
List,
I have an 8.1.7.4 OpenVMS server that I export some of the tables
from and import them into a RedHat AS2.1u3 9.2.0.5 database. I was
getting the IMP-00020 (long column to large for column buffer size)
error when I tried the import today.
Researching Metalink and trying thier resolutions did not work and I
still got the errors. DIRECT,RECORDLENGTH,BUFFER,COMMIT options of
different combinations still failed.
I have done the export/import combinations on these Oracle versions in
the past and all was successful.
I ftp'd the files from the OpenVMS server directly to the RedHat server
using the BINARY mode.
Reading the Metalink forum I noticed ONE small note that a user used to
sovle his problem...he ftp'd the files from the server to his pc and
then from the pc to another server...PROBLEM SOLVED. This is the normal
method I use but today I am trying to eliminate the pc in the middle.
I used the server - pc - server method and all is well.
Ron
 |
 |