Hello,
I'm in the process of doing the same:
The idea of a tablespace in maxdb does not exist -> so don't worry about
that. Just create your datafile(s)
Memoryspace -> take the default proposed during db creation, I will do
tuning afterwards. But the defaults seem to do very well
Logical en physical space -> there is no seperation in maxdb. If you
want to emulate schema like structures: this is what I have done:
suppose you have schema A,B,C; than in maxdb
1) log in with the sysdba user account
2) Create 3 different users (A,B,C) of type resource (create user A
password x resource not exclusive).
3) Log out of sysdba and log in as user A. Create all the
tables,sequences,triggers, ... that where in schema A now in user A. Do
the same for user/schema B and C
4) If you want user B to have access to the tables of user A, grant them
access. Then user B can access the other tables with the familiar select
* from A.tablename syntax.
Diskspace -> don't know what you mean
How to make the export -> creating perl scripts that connect to both
oracle and maxdb using ODBC seems to be the best solution. I started
with using MSAccess :-$ . I linked to both databases and than create SQL
queries a la "INSERT INTO MAXDB_A.table SELECT * FROM SCHEMA_A.table"
But this only works for small tables.
Maybe some other things I learned:
In oracle 9 it is possible to generated DDL statements of all the
objects. Can be very handy with some post processing/find_and_replace
ORACLE number(n) in MAXDB 0<n<=5 -> smallint (=fixed(5,0))
6<=n<=10 -> int(eger) (=fixed(10,0))
n>10 -> fixed (p,s)
p = number of digits, s= number of decimals / when s is not defined it
defaults to 0
ORACLE float in MAXDB fixed
float(1<=x<=38)
ORACLE clob in MAXDB varchar(8000) if type is asci, if type is unicode
it cannot be longer than 4000. The maximum overall rowlength is 8088
long (asci | unicode) Long is not comparable nor searcheable .
Total
length is 2GB
ORACLE blob in MAXDB long(byte)
The general syntax for table creation is:
CREATE TABLE <table_name> (<column_name> <data_type>(<column_attrib>)
[NOT NULL], ...
PRIMARY KEY (<column_name>,...),
FOREIGN KEY [<referential_constraint_name>] (<referencing_column>)
REFERENCES <referenced_table> [(<referenced_column>,...)],
CONSTRAINT <constraint_name> CHECK <search_condition>,
CONSTRAINT <constraint_name> UNIQUE (<column_name>,...)
COMMENT ON <table_name>.<column_name> IS 'the comment you want to have'
If you want to do batch scripts; seperate the commands in your script
with '//'
If you use these function in your oracle queries, in max db they are as
follow:
to_number(string) => num(string)
to_char(number) => chr(integer)
to_date(string) => char(string)
And don't forget: "order by on a subselect is not possible in maxdb" I
learned that the hard way.
I started writing a documentation on everything I encountered during my
migration. I hope to find some time to finish a draft in the near
future.
Regards and good luck
Filip Sergeys
On Fri, 2004-01-09 at 18:43, [EMAIL PROTECTED] wrote:
> We want to export a database from Oracle and we need to know informations about :
> - Tablespace
> - Memoryspace
> - Logical and physical space
> - Diskspace
> - How to make the export
> Thanks
>
>
> --
> MaxDB Discussion Mailing List
> For list archives: http://lists.mysql.com/maxdb
> To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
>
--
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
* System Engineer, Verzekeringen NV *
* www.verzekeringen.be *
* Oostkaai 23 B-2170 Merksem *
* 03/6416673 - 0477/340942 *
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
--
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]