Database replication

buddies,

I hava a scenerio where i have replicated the database/table from IBM informix to Orcale 10g using some migration tool.

Now i got to verify or validate whether the replicated data in destination are same as that of source databse or corrupted.

Of course, i can make use of count(*). But My tables contain very huge data.

I heard that it is possible to verify using Checksum or hash.. and also API's in java are available such as DBReplicator and sync4j.

How to validate them.? Could anyone help me out to resolve this issue?

could anyone please answer my question?

There are plenty of database diffing tools out there (for example this is one). I'd try something like that.

Thanks for your reply..

could you give me any tutorial for using any of such tools? (it should be for heterogeneous databases)

I am not able to find any tools from the link you specified.

Could you help me out on this?

is there difference between checksum in oracle and checksum in informix?

It will be great if i am able to find the checksum for the table in both Oracle and informix. Will this ( i.e. checksum) be useful to find if the replicated DB is consistent with source DB, in the heterogenous environment?

please anyone answer my questions ASAP..

Thanks in advance.

Sorry, here's a different link: database diff.

could you give me any tutorial for using any of such tools? (it should be for heterogeneous databases)

No, but there is sure to be plenty of documentation available for any of these tools. I'd start by reading that (after you choose which one you want to use of course).

ok..

is there any significant difference checksum ( for table ) in oracle and informix?

Coulcd anyone answer my question?

I'm not sure how you generate a checksum for an entire table in Oracle, its not something I've heard of people doing. I know Oracle can use checksums to compare data blocks, but I can't think how you could use this and how you could compare anything other than another Oracle database anyway. Any reason you want to use a checksum rather than a database diffing tool?

ok. thanks for your reply..

My Problem:
After performing replication on the heterogeneous databases (done only once for application migration), I should check the consistency in the database. I need to verify if all the columns and rows are not corrupted.

Possible solution:
Here i need to use the generic solution for few heterogeneous databases to verify and validate the data in the databases.
Is there any algorithm (in java ) to achieve this ? It could be using checksum or any other tactics..

could you help me out on this?

Pual,

Could you suggest me any optimal solution for the problem i explained above?

I think I already have. It's what I'd do anyway.

Thanks.