SDK Support for Bulk Transfers - DDD and Event Sourcing Design

Design overview

This diagram gives an overview of the SDK design.

Design Overview

An http 202 response when posting a asynchronous request means that the the SDK has accepted the request, and that the request will be processed and a response provided. Because of the potential long delay's involved in processing large numbers of bulk payments in an asynchronous way. A new design approach in the SDK was necessary to meet the 202 response expectations.

DDD and Event sourcing design

An event sourcing and domain driven was chosen, as this solves all requirements of reliability and scalability while takeing advantage of the lib's and tools that have been built for the Mojaloop vNext architecture.

Bulk Payer DFSP SDK-Scheme-Adapter

Outbound event sourcing sequence diagram

Outbound sequence diagram

Payee DFSP SDK-Scheme-Adapter

Inbound bulk quotes event sourcing sequence diagram

Inbound bulk quotes sequence

Inbound bulk transfers event sourcing sequence diagram

Inbound bulk transfers sequence

Redis Store Data mapping for outbound bulk transfer

1. States (Global and individual)

Command:

HSET <key> <attribute1> <value1>

Key:

outboundBulkTransaction_< bulkTransactionId >

Attributes:

• bulkTransactionId: bulkTransactionId

• bulkHomeTransactionID: Home transaction ID

• request: { options: Options, extensionList: Bulk Extension List }

• individualItem_< transactionId >: Serialize ({ id: transactionId request: {} state: Individual state batchId: <UUID> partyRequest: {} quotesRequest: {} transfersRequest: {} partyResponse: {} quotesResponse: {} transfersResponse: {} lastError: {} acceptParty: bool acceptQuotes: bool })

• state: Global state

  • RECEIVED

  • DISCOVERY_PROCESSING

• bulkBatch_< batchId >: Serialize ({ id: batchId state: Individual state

  • AGREEMENT_PROCESSING

  • TRANSFER_PROCESSING bulkQuoteId: <UUID> bulkTransferId: <UUID> (Can be batchId) })

• partyLookupTotalCount: Total number of party lookup requests

• partyLookupSuccessCount: Total number of party lookup requests those are succeeded

• partyLookupFailedCount: Total number of party lookup requests those are failed

• bulkQuotesTotalCount: Total number of bulk quotes requests

• bulkQuotesSuccessCount: Total number of quotes requests those are succeeded

• bulkQuotesFailedCount: Total number of quotes requests those are failed

• bulkTransfersTotalCount: Total number of bulk transfers requests

• bulkTransfersSuccessCount: Total number of bulk transfers requests those are succeeded

• bulkTransfersFailedCount: Total number of bulk transfers requests those are failed

::: tip Notes

• Kafka messages should contain bulkID.

• To update the global state use the command HSET bulkTransaction_< bulkTransactionId > state < stateValue > :::

2. For mapping individual callbacks with individual bulk items

Command:

HSET outboundBulkCorrelationMap <attribute1> <value1>

Attributes:

• partyLookup_<id_type><id_value>(<subid_type>): "{ bulkTransactionId: <bulkTransactionId>, transactionId: <transactionId> }"

• bulkQuotes_<bulkQuoteId>: "{ bulkTransactionId: <bulkTransactionId>, batchId: <batchId> }"

• bulkTransfers_<bulkTransferId>: "{ bulkTransactionId: <bulkTransactionId>, batchId: <batchId>, bulkQuoteId: <bulkQuoteId> }"

• bulkHomeTransactionId_<bulkHomeTransactionId>: "{ bulkTransactionId: <bulkTransactionId> }"

::: tip Notes:

• We can use HKEYS command to fetch all the individual transfer IDs in a bulk to iterate :::

Redis message format for inbound bulk transfer

1. Bulk Quotes

Command:

HSET <key> <attribute1> <value1>

Key:

inboundBulkQuotes_< bulkQuotesId >

Attributes:

• bulkQuotesId: bulkQuotesId

• individualItem_< quotesId >: Serialize ({ id: quotesId request: {} state: Individual state quotesRequest: {} quotesResponse: {} lastError: {} })

• state: Global state

  • RECEIVED

  • PROCESSING

• bulkQuotesTotalCount: Total number of bulk quotes requests

• bulkQuotesSuccessCount: Total number of quotes requests those are succeeded

• bulkQuotesFailedCount: Total number of quotes requests those are failed

::: tip Notes

• Kafka messages should contain bulkQuotesId.

• To update the global state use the command HSET bulkQuotes_< bulkQuotesId > state < stateValue > :::

2. Bulk Transfers

Command:

HSET <key> <attribute1> <value1>

Key:

inboundBulkTransfer_< bulkTransferId >

Attributes:

• bulkTransferId: bulkTransferId

• individualItem_< transferId >: Serialize ({ id: transferId request: {} state: Individual state transfersRequest: {} transfersResponse: {} lastError: {} })

• state: Global state

  • RECEIVED

  • PROCESSING

• bulkTransferTotalCount: Total number of bulk transfers requests

• bulkTransferSuccessCount: Total number of transfers requests those are succeeded

• bulkTransferFailedCount: Total number of transfers requests those are failed

::: tip Notes

• Kafka messages should contain bulkTransferId.

• To update the global state use the command HSET bulkTransfer_< bulkTransferId > state < stateValue > :::

3. For mapping individual callbacks with individual bulk items

Command:

HSET inboundBulkCorrelationMap <attribute1> <value1>

Attributes:

• quotes_<quoteId>: "{ bulkQuoteId: <bulkQuoteId> }"

• transfers_<transferId>: "{ bulkTransferId: <bulkTransferId>, bulkQuoteId: <bulkQuoteId> }"

::: tip Notes:

• We can use HKEYS command to fetch all the individual transfer IDs in a bulk to iterate :::