The CData ODBC Driver for Amazon DynamoDB offers bulk load support in Amazon DynamoDB through the ODBC Batch API. The driver can execute related SQL data manipulation statements simultaneously by translating them into a single bulk API request. In your application's code, the driver executes statements based on an array of inputs and a parameterized query.

Batch Update Procedure

Complete the following steps to execute a batch update:

1. Define arrays of column values for each parameter in the statement.
2. Set the SQL_ATTR_PARAMSET_SIZE statement attribute.
3. Bind each array to each parameter.
4. Execute the parameterized statement.

Bulk Insert

The following code shows how to execute a bulk insert with SQLSetStmtAttr, SQLBindParameter, and SQLExecDirect:

SQLHSTMT pHstmt = NULL;
const int MAX_INSERT_COUNT = 2;
const int MAX_BUFFER_SIZE = 100;
char NameArray[MAX_INSERT_COUNT][MAX_BUFFER_SIZE] = { 0 };
SQLLEN cbNameArray[MAX_INSERT_COUNT] = { 0 };
...
retcode = SQLSetStmtAttr(pHstmt, SQL_ATTR_PARAMSET_SIZE, (SQLPOINTER)MAX_INSERT_COUNT, 0);
char *Name1 = "Jon Doe", *Name2 = "John";
cbNameArray[0] = strlen(Name1);
cbNameArray[1] = strlen(Name2);
memcpy(NameArray[0], Name1, strlen(Name1));
memcpy(NameArray[1], Name2, strlen(Name2));
retcode = SQLBindParameter(pHstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 0, 0, NameArray, MAX_BUFFER_SIZE, cbNameArray);

retcode = SQLExecDirect(pHstmt, (SQLCHAR*) "INSERT INTO Account (Name) VALUES (?)", SQL_NTS); 

To retrieve the new records' Ids, query the LastResultInfo table:

SQLExecDirect(hstmt, (SQLTCHAR*)"SELECT * FROM LastResultInfo#TEMP", SQL_NTS);  
SQLTCHAR sId[20] = {0};
SQLLEN cbId = 0;
SQLBindCol(hstmt, 1, SQL_C_CHAR, sId, 20, &cbId);
SQLRETURN retcode = SQLFetch(hstmt);   

Bulk Update

An array of the primary keys of the records to update must be specified. The following code shows how to execute a bulk update with SQLSetStmtAttr, SQLBindParameter, and SQLExecDirect:

SQLHSTMT pHstmt = NULL;
const int MAX_UPDATE_COUNT = 2;
const int MAX_BUFFER_SIZE = 100;
char IdArray[MAX_UPDATE_COUNT][MAX_BUFFER_SIZE] = { 0 };
char IdArray[MAX_UPDATE_COUNT][MAX_BUFFER_SIZE] = { 0 };
SQLLEN cbIdArray[MAX_UPDATE_COUNT] = { 0 };
SQLLEN cbIdArray[MAX_UPDATE_COUNT] = { 0 };
...
retcode = SQLSetStmtAttr(pHstmt, SQL_ATTR_PARAMSET_SIZE, (SQLPOINTER)MAX_UPDATE_COUNT, 0);
char *Name1 = "Jon Doe", *Name2 = "John";
char *Id1 = "Id1", *Id2 = "Id2";
cbNameArray[0] = strlen(Name1);
cbNameArray[1] = strlen(Name2);
cbIdArray[0] = strlen(Id1);
cbIdArray[1] = strlen(Id2);
memcpy(NameArray[0], Name1, strlen(Name1));
memcpy(NameArray[1], Name2, strlen(Name2));
memcpy(IdArray[0], Id1, strlen(Id1));
memcpy(IdArray[1], Id2, strlen(Id2));
retcode = SQLBindParameter(pHstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 0, 0, NameArray, MAX_BUFFER_SIZE, cbNameArray);
retcode = SQLBindParameter(pHstmt, 2, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 0, 0, IdArray, MAX_BUFFER_SIZE, cbIdArray);

retcode = SQLExecDirect(pHstmt, (SQLCHAR*) "UPDATE Account SET Name = ? WHERE Id = ?", SQL_NTS); 

Bulk Delete

An array of the primary keys of the records to delete must be specified. The following code shows how to execute a bulk delete with SQLSetStmtAttr, SQLBindParameter, and SQLExecDirect:

 
SQLHSTMT pHstmt = NULL;
const int MAX_DELETE_COUNT = 2;
const int MAX_BUFFER_SIZE = 100;
char IdArray[MAX_DELETE_COUNT][MAX_BUFFER_SIZE] = { 0 };
SQLLEN cbIdArray[MAX_DELETE_COUNT] = { 0 };
...
retcode = SQLSetStmtAttr(pHstmt, SQL_ATTR_PARAMSET_SIZE, (SQLPOINTER)MAX_DELETE_COUNT, 0);
char *Id1 = "Id1", *Id2 = "Id2";
cbIdArray[0] = strlen(Id1);
cbIdArray[1] = strlen(Id2);
memcpy(IdArray[0], Id1, strlen(Id1));
memcpy(IdArray[1], Id2, strlen(Id2));
retcode = SQLBindParameter(pHstmt, 1, SQL_PARAM_INPUT, SQL_C_CHAR, SQL_CHAR, 0, 0, IdArray, MAX_BUFFER_SIZE, cbIdArray);

retcode = SQLExecDirect(pHstmt, (SQLCHAR*) "DELETE FROM Account WHERE Id = ?", SQL_NTS);