E1812 Unicode: How to make Faster

JDE suggest that we can put tables in different datasource and run Unicode for both DataSource and then merge the two data Source. This looks me too quite unhandy.

The one way i tested is :-

Dividing UniDataConv.xml in 2 halves and running both of them simultanesously from 2 machines.

This works for set of few tables. 10 - 10 Tables in each xml file.

Could be handy way.
 
Hi ALL,

I just came across a question in a mind and would like to share with you.

Data Type: - The types of data like CHAR, VARCHAR, NUMBER, NCHAR etc.

Data: Actual Data in particular language. Assume and lets talk about English.

Now I have two tables assume F00023 and F0091 in Non Unicode Format (XE Format)

Both tables in XE have Data types of CHAR, NUMBER, VARCHAR etc.

F00023 have 0 records.
F0911 have millions record.

Assume all activities installing 812, create upgrade plan, data conversion for 812 all done.
And we have to do now Unicode.

During Unicode process of JDE, i kept close eye and find that it first copy the table as F00023_NONUNI converts the original table F00023 DATA TYPES in UNICODE Format and then copy Data from F00023_NONUNI to F00023 table. This occurs quickly as there is no data in table.

Now comes F0911 Table.
Copy Original Table F0911 to F0911_NONUNI. (Tooks 3 –5 hours for 10 –50 millions of record)
Create empty table F0911 with DATA TYPES as NCHAR, NVARCHAR etc (i.e. in Unicode format)
and Then copy data from F0911_NONUNI to F0911 tables. (again 3 – 5 hours)
And then making run for each index with millions of record in table makes around 10 –15 hours run for a single table like F0911.


If the setup have other big tables like F0911 then usually it tooks long time to Run Unicode.

The quick questions are:-

1. Do Unicode converts the data also? I feel it don’t convert data into Unicode format. Its just Convert DATA Types new Unicode format.

2.In regard to save Unicode Conversion time for big table, can we took a back of table of Tables like F0911 run unicode for F0911 and then manually insert data back.
Export and Import Dump on F0911 are much quicker then JDE insert data?
 
[ QUOTE ]
1. Do Unicode converts the data also? I feel it don’t convert data into Unicode format. Its just Convert DATA Types new Unicode format.

[/ QUOTE ]

The unicode conversion program only converts the metadata, it doesn't convert the data itself.

[ QUOTE ]
2. In regard to save Unicode Conversion time for big table, can we took a back of table of Tables like F0911 run unicode for F0911 and then manually insert data back.
Export and Import Dump on F0911 are much quicker then JDE insert data?

[/ QUOTE ]


The process can be customized to your needs, and I know of at least one JDE customer who has performed the entire conversion to Unicode outside of the canned tool. You won't be able to get support from Oracle if you choose to go that route, but Oracle and/or independent consultants could certainly support you.
 
Back
Top