How to commit each DB transaction in mapping node in IIB Toolkit? - ibm-integration-bus

Question

I have a JSON body of Array type. Now each body will be inserted into DB one by one. There is constraint to use mapping node only as it is having GUI. But in mapping node, I cannot change the transaction from automatic to any other value. As a result when first body is inserted into DB and second one is failed, first one is also getting rolled back. How to design the mapping node such a way that when 1st is successful, it will be committed too? or if the 1st is failed, second will also be attempted to be inserted?
here is the JSON request of atleast 2 elements' array:
[
{
"batch_id": 2019,
"trx_no": "test",
"sys_ref_no": "test",
"trx_line_num": 1,
"trx_start_date": "test",
"trx_post_on_date": "test",
"trx_post_on_date_dr": "06-APR-20",
"company_code": "test",
"company_id": 1,
"shop_num": "test",
"shop_id": 1,
"pos_num": "test",
"pos_id": 1,
"tax_type": "test",
"tax_code": "test",
"tax_rate": 1,
"currency_code": "test",
"currency_id": 1,
"tax_amt": 1,
"tax_id": 1,
"tax_base_amt": 1,
"tax_exchange_rate": 1,
"tax_exchange_date": "test",
"tax_exchange_type": "test",
"status": "test",
"error_message": "test",
"creation_date": "06-APR-20",
"created_by": 1,
"last_update_date": "06-APR-20",
"last_update_by": 1,
"data_source": "test",
"request_id": 1,
"billing_intf_flag": "test",
"n_attribute1": 1,
"n_attribute2": 1,
"n_attribute3": 1,
"c_attribute1": "test",
"c_attribute2": "test",
"c_attribute3": "test",
"d_attribute1": "test",
"d_attribute2": "test",
"d_attribute3": "test",
"d_attribute1_dr": "06-APR-20",
"d_attribute2_dr": "06-APR-20",
"d_attribute3_dr": "06-APR-20"
},
{
"batch_id": 2020,
"trx_no": "test",
"sys_ref_no": "test",
"trx_line_num": 1,
"trx_start_date": "test",
"trx_post_on_date": "test",
"trx_post_on_date_dr": "06-APR-20",
"company_code": "test",
"company_id": 1,
"shop_num": "test",
"shop_id": 1,
"pos_num": "test",
"pos_id": 1,
"tax_type": "test",
"tax_code": "test",
"tax_rate": 1,
"currency_code": "test",
"currency_id": 1,
"tax_amt": 1,
"tax_id": 1,
"tax_base_amt": 1,
"tax_exchange_rate": 1,
"tax_exchange_date": "test",
"tax_exchange_type": "test",
"status": "test",
"error_message": "test",
"creation_date": "06-APR-20",
"created_by": 1,
"last_update_date": "06-APR-20",
"last_update_by": 1,
"data_source": "test",
"request_id": 1,
"billing_intf_flag": "test",
"n_attribute1": 1,
"n_attribute2": 1,
"n_attribute3": 1,
"c_attribute1": "test",
"c_attribute2": "test",
"c_attribute3": "test",
"d_attribute1": "test",
"d_attribute2": "test",
"d_attribute3": "test",
"d_attribute1_dr": "06-APR-20",
"d_attribute2_dr": "06-APR-20",
"d_attribute3_dr": "06-APR-20"
}
]

Answer 1

There is constraint to use mapping node only as it is having GUI.
This problem would be simple to solve using a Compute node and a small amount of ESQL. It is (as far as I know) impossible using the Mapping node. So either that constraint gets relaxed, or the requirement for independent transactions has to be removed.
In general, it is a bad idea to make rigid rules about which IIB transformation language to use, because it leads to awkward situations like this one. It's OK to have a preferred language or a default language, as long as other languages are allowed when it makes sense. Your situation is actually slightly unusual. It is more common (but still unwise) for ESQL to be the one-and-only-allowed-language!

Answer 2

Since don't want to process this JSON as single unit of work i.e. either commit to db or rollback, you could divide the work into two separate message flows. The first message flow splits the input JSON with multiple childs into individual JSON with only one node. Feed this individual JSON node(body) to the next message flow which uses the mapping node to insert in the database. This way you can workaround your constraint of using mapping node.