Integrating Blockchain with the Mainframe

blockchain-3750157_960_720.jpg

Blockchain strives to eliminate the need for middlemen by building trust between business parties, automating certain transaction processes, securing transaction data, and reducing costs across the board by creating an immutable ledger or log of transactions or data. Blockchain can simplify the way business is conducted, but it is not a replacement for the database systems in place today. A database is a system to store data, but a blockchain is a ledger to record and log movements.[1] Unlike a traditional database, on a blockchain each participant maintains, calculates, and updates new entries into the database, and all nodes work together to ensure they reach the same conclusions. This provides in-built security for the network without requiring the access and permissions required by a central administrator in charge of securing a database.[2]

According to a Stanford University study, businesses can improve the trust in digital transactions by using digital signatures and blockchain (distributed ledger) technology to provide a secure and verifiable means to determine asset ownership and data provenance. In essence, blockchain becomes a new model for transactions that cross traditional business boundaries, ranging from hard goods and financial assets, to intangibles like patents, to record what happens between businesses.

In his “Blockchain Demystified” presentation at SHARE St. Louis, IBM Executive IT Architect Wilhelm Mild notes that in current real estate transactions, the real estate firm could have a WAS web server running JAVA applications and IMS apps on an IBM Z mainframe, and the bank could have a CICS system running on a different mainframe, both of which log application information differently. Each of these firms also has a separate ledger for each transaction with its business partners that collects the data relevant to each firm, but often the data collected on those separate ledgers is not the same. To ensure the transaction process is trusted on both sides, there are middlemen (such as central banks), which are trusted by both parties to ensure transactions are accurate and valid.

To interface with a blockchain from an existing system (mainframe or other), businesses either need to create APIs through a software developer kit or locate an existing API to communicate with the blockchain and record transactional data as agreed upon within the peer network. APIs can connect over channels to one or more peers in the blockchain, connect over one or more channels to the orderer nodes, receive events from peers on the blockchain, or receive client crypto material from a local managed service provider. Businesses also can connect current databases and systems of record with multiple blockchains, such as a blockchain with the bank for financial transactions and a blockchain with a supplier, to ensure the correct transactions are made at the right time.

The Blockchain Insurance Industry Initiative (B3i), for example, involves a consortium of more than 15 insurers and reinsurers who will use blockchain to create a platform through which ceding, handling, and trading risks will become more efficient for all participants. Typically, insurance contracts can vary in policy language and the claims process can take a long time to close due to disagreements in policy language or the slow movement of paperwork. B3i is expected to eliminate the disparities in claims processing by eliminating the back and forth between reinsurers and insurers until they agree on claims. Smart contracts and any updates on the blockchain would be viewable by any participating insurer or reinsurer. This would prevent disputes similar to the $3.4 billion disagreement over which insurance policy language was the prevailing coverage when two planes crashed into the World Trade Center in 2001 and would determine whether the crashes were one event or two.[3]

Mild adds that businesses can integrate blockchain through business process management, operational decision manager, IBM Integration Bus, and z/OS Connect EE. Many blockchain users rely on a loopback server to generate a REST API, from which a Swagger page can be generated to connect to z/OS Connect. This enables the mainframe to communicate with blockchain through REST. CICS or IMS and the client application will use REST to talk to blockchain when transactions based on the smart contract parameters occur. Smart contracts also can generate events, such as verifying a transaction was recorded to the blockchain. Because CICS and IMS process transactions quickly, the systems are not going to wait for a blockchain response, but the system can record data to the blockchain ledger to enable the smart contract to generate an event and communicate to CICS or IMS that the transaction was verified as complete.

In a joint “Integrating Your System with Blockchain” presentation with Mild, IBM Integration Architect Jennifer Foley says that mobile client apps must be able to communicate with the mainframe or REST server on the blockchain directly, and the information generated from the communication via blockchain is the data that the CICS system, for example, needs to record in its own ledgers using COBOL or whatever language the system uses to record transactions. She explains that an insurer, for instance, can call on the blockchain for leased car details, including VIN numbers, before writing its policy on the car. Using a build-out kit to create an API copybook from Swagger to translate data from blockchain into COBOL artifacts, for example, ensures that not only the VIN number of the car but also its other details, such as make and model, can be communicated to CICS and recorded on the backend, Foley says. Additionally, companies may want to expose assets on the mainframe as REST APIs with the help of API mapping, so peers on the blockchain can make calls to CICS or IMS, creating a two-way communication with the blockchain.

Hosting a blockchain on a mainframe, says Christopher Tozzi in a Syncsort blog post, can not only double down on data reliability but also take advantage of the computing power of the mainframe itself. Blockchains have compute-intensive tasks, such as cryptographic hashing, and commodity servers, for example, are not always up to the challenge. Mainframes, he adds, also could help close an important infrastructure gap for companies seeking to build their own blockchains.

Blockchain enhances trust between business partners and becomes an auditable source of data, should issues arise in a transaction or supply chain. These immutable or tamper-proof ledgers become a single source of truth, according to Mild. Businesses of the future could exchange assets via blockchain smoothly, reliably, and without the need for an intermediary, as they engage one another through different channels on blockchains to move assets across the globe. Reducing the need for a middleman will ultimately reduce transaction costs and increase efficiency throughout the supply chain, financial transaction process, and more.

Learn more about blockchain in from this SHARE’d Intelligence article: Blockchain 101: How Mainframes Can Power Blockchain for Businesses, from SHARE’d Intelligence.

 

[1] https://medium.com/@chainfrog/5-reasons-that-blockchain-is-not-just-a-slow-database-55fe9d913578

[2] https://www.coindesk.com/information/what-is-the-difference-blockchain-and-database

[3] https://www.reuters.com/article/us-insurance-blockchain-swiss-re-idUSKBN15924K

Recent Stories
SHARE Phoenix: Women in IT Initiative Makes a Splash

SHARE: GAO’s Mainframe Risk Claims Debunked

Lost in Translation