Hi, all
I'm having a problem in creating table from a sequential dump file. I have dump the entire tables which resulted a huge file size of 5.6GB and then tried to create back the table and the message "wrong dumpfile".
(Baan IVc, AIX 4.3, Informix 7.31). Pls advice.

TQ

Note : No problem with single table dumps.

Before you do anything keep a backup of your dumpfile.

1) Please check the separator used to create the dump file.

2) The first two lines of your dump file are crucial, I hope anyone has not played with them. If yes, try creating another dump in the same manner and use the first two lines of this new file.

1. did not use the separator option just standard dump all the tables into one file.

2. Check the header, it is the same.

Rdgs

Maybe you have a problem with the file size. Try to split dump to more files with –M option.

I see 40c2/AIX4.2 on your profile but the post mentions 40c4/AIX4.3
Are you copying data between different systems?
There were improvements done on a portingset that caused those types of messages, you need to be sure that the portingsets of the different systems support the same dump format.

Thanks everyone for your advice,
I've resorted to dump using the field separator which resulted in dumping every individual file into different file each and then upload back using the same format.
Completed

Rdgs

Can any body tell me what is the difference between field separator and standard dump of all tables in one file????
Its very Urgent!!!!!!!!!!!!!

Field seperator will have 1 record per line with your specified field seprator
it will also have header info:
2006-01-13 12:22:55.707
#$!pre-dump!$#00037cv

and has a .S extention.

A standard dump of a table will:
1. start out with a header:
2006-01-13 12:23:35.438
#$!pre-dump!$#000300v
#$dttaad200

2. List the table DD

3. List each record with a null '\0' seperator

Can't imagine why you'd need to know that urgently. but i hope the world is saved...

Dave

Thanks Dave

Can i able to import multibyte characters by using pipe ("|") sepeartor in "Create tables from sequential dump" session
or
shall I use standard dump of table??

it should not matter - old porting sets had problems with seperators
but newer ones work fine (so if this is LN you should be fine either way)

I assume your baan environment is setup for Multibyte?

Dave

Thanks Dave.

I have one more query!!!!!!!!!!

I am using LN..
and i am exporting data from 500, 501 compnay and want to import 600 and 601 compnay..
Now, if I used standard dump of table for "Create tables from sequential dump" session, and now I want to change company number froms in company number fields, how shall I achive same?
As I am using standard dump of table which is not a field separtor, so there will be problem with using "copy company data" session and also in LN tfgld9004m000 session are not available.

and because of above problem I am thinking to use "|" separator or sequential dump for "Create tables from sequential dump" session for my multibyte charaters..

What you suggest??

Changin the company # itself is easy (just use -c option to bdbpost)
but changing the internal table references to the company requires
you to use the change company # session which requires a separator
i believe. So you'll have to re-dump your file and use a seprator.

Dave

But dave in that case, I have to use "|" separator, will that import my multibyte charaters??

Yes. you don't have to use "|" though, you can use whatever you want.

Dave