Binary Dump Code Page

[vincenteam]

Hi,

I'm migrating fron Progress v9.1E 02 to OpenEdge 10.1B 02 betewenn two SUN servers.

The V9 database has ibm850 codepage and the OE 10 db has iso8859-1 codepage.
The question is: can we run the binary dump process with iso8859-1 codepage or run the binary load process with ibm850 codepage ?

Thanks in advance.

 

[alex10]

Hello all, I have this same question, anyone can help us?

Thanks,

 

[Casper]

You can't. Either convert the ibm-850 database first to iso8859-1 or convert the 10.1B database after load to iso8859-1.

From the documentation:

PROUTIL DUMP does not record the code page of the data it writes to the dump file.
Dumped data can only be loaded to a database with the same code page as the source
database. Use the ASCII dump and load utilities to perform a code page conversion with​

your dump and load

BTW 10.2A is the latest release,so while you are at it, use this one

Regards,
Casper

 

[alex10]

Thanks Casper,

I verify and both my databases has the same code page, now when i try the binary load it prompt for a confirmation, is possible to set Yes as default answer? because i have a script to binary load

hp ux 11.v3
old: OE 9.1e
new: OE 10.2A

/pro102a/dlc/bin/proutil /mydb/db10/db1 -C load /mydb/db10/tmp/db1/usrc_ctrl.bd
Warning: input .bd file contains no code page information.
If you are certain the .bd file originated from a database with code page iso8859-1 then enter Y to continue. (10853)

 

[Casper]

If you are sure that both databses have the same codepage then you can enter yes.
I never encountered this question during a dump/load. Are you really sure the database has not the undefined codepage?
Did you perfrom the binary dump yourself?

Casper.

 

[alex10]

Yes, for the dump i use:

/pro91e/dlc/bin/proutil /mydb/dbprod/db1 -C dump usrc_ctrl /mydb/db10/tmp/db1

for the load

/pro102a/dlc/bin/proutil /mydb/db10/db1 -C load /mydb/db10/tmp/db1/usrc_ctrl.bd

and then i get
Warning: input .bd file contains no code page information.
If you are certain the .bd file originated from a database with code page iso8859-1 then enter Y to continue. (10853)

If i answerd Y it load without a problem, but there are many tables and i dont want do it manually.

Thanks again

 

[RealHeavyDude]

Just one note:

ibm850 and iso8859-1 contain the same characters but on different positions ( for example the German Umlauts ). Therefore, if you load data that was dumped binary from a ibm850 database into one with iso8859-1 you will find these characters screwed up. Please be aware that iso8859-1 like ibm850 do not contain the € sign!

Since both contain the same characters ( each character of one code page is present in the other one ) you can easily change the code page of either the source databaes or the target database so that the code pages match ( proutil db-name -C convchar convert ).

In this special case clients that use one of the code pages can connect to the database seamlessly ( Progress will do the code page conversion for you without you noticing ).

HTH, RealHeavyDude.