Compare two datatable and find which record is modified/newly addded/No change

I would sort each table by one of the fields, the key field, and then walk down both tables

While (counterA < tableARecords && counterB < tableRecords)<br />
{<br />
  if (listA[counterA].Index == listA[CounterB].Index)<br />
  {<br />
    // compare records<br />
    counterA++;<br />
    counterB++<br />
  }<br />
  else if (listA[counterA].Index > listA[CounterB].Index)<br />
  {<br />
    //This means that a record in listA is not in listB<br />
    counterB++;<br />
  }<br />
  else<br />
  {<br />
   //This means that a record in listB is not in listA<br />
    counterA++;<br />
  }<br />
}<br />
//Deal with extra records in listA or listB here

The simple way to achieve that doesn't exist, a specially when more than one user can add, edit and delete data in the MS Excel file.
To compare data, you need to loop through all records in both tables (see solution 2 and 3).

If the count of records in your master table is less than few thousands, you can compare data on Excel level. Why?
1) To set ID for each equal record in MS Excel file and add "comparision date", and to eliminate it from future comparisions,
2) To improve future comparisions:
a) when user made changes in the existing record (with ID), need to clear "comparision date", to recognize it as "after changes",
b) when user added new record, should leave ID and "comparision date" field empty, to recognize it as a new record.
The rest of data are new or changed. What can i say? Hand-made work is waiting on you... ;(
Without knowing the structure of both tables and relationships between data we can't help you more...

The best way to avoid trouble is to write a program to communicate with the Oracle server.

When you're wondering about looping logic, it's best to think about how you would do this if you had to work by hand with paper and pencil.

You'd start with the first master record. You'd run your check on that record. Then you'll need to check each of the child records that correspond to it. Once you're done you'll move on to the second master record.

I imagine your loops would look soemthing like this:

'Loop the Master Table
For Each rowMaster as DataRow in dtMaster.Rows
   'Insert your code here that checks if the master row has been updated

   'Loop through this master record's child records
   For Each rowChild as DataRow in dtChild.Select("MasterIdFromChildTable=" & rowMaster("MasterIdFromMasterTable"))
      'Insert your code here that checks if the child row has been added
   Next
Next

You may have to arrange it so that any updating you do the Master table is done after all of the looping. The For Each may not work if you add/delete rows to the table you are looping and you may get an exception. Also, it will work faster if you wait to update everything at the end.

Hope this helps get you started.