Distributed Computing with Permissioned Blockchains and Databases

Summary

The topic of blockchains, and in particular that of permissioned blockchains, has rapidly gained interest in both the industrial and the research communities in recent years. It particularly pertains to situations where trust among several parties that are about to do business together is difficult to establish (e.g., due to organizational, financial, or timing reasons) or impossible to establish at all. A blockchain is a decentralized, distributed ledger that consists of immutable blocks containing transactions that can be accessed by any party, and that provides trust via replication over all nodes and an agreed-upon execution order of the transactions. Of particular interest are permissioned blockchains in which the associated parties are known and authenticated, yet still do not fully trust each other.

Many applications have shown interest in the concept of blockchains, since the situation just described applies to many real-world scenarios, including (global) supply chains, the Internet of Things, connected cars, manufacturing, banking, and healthcare. As a consequence, a number of players in the IT industry work on a development of the technology, and several consortia have been formed to advance the technology across industries, among them Hyperledger and R3. Moreover, a number of companies have released Blockchain-as-a-Service (BaaS) platforms, including IBM, Oracle, Amazon, Baidu, and Alibaba.

The technology has many links into the database community; however, the situation is basically like it was in the database area many years ago, when only a few systems had been released but users were on their own to figure out how to use them effectively. As the seminar has shown, many interesting issues remain to be solved, and there is a wide variety of aspects and research issues currently under investigation. Of these, the following were discussed:

• Blockchain scalability w.r.t. transaction throughput, one of the main roadblocks to business adoption
• Transaction ordering and endorsement, consensus of transaction commit
• AAdjustments to the Proof of Work (PoW) consensus mechanism, other optimizations to consensus algorithms (e.g., Byzantine consensus) in the presence of transaction failures and in light of scalability
• Block validation
• Languages for smart-contract specification (e.g., Sandcastle SQL and Solidity)
• Amendments to Hyperledger Fabric, such as channels
• Cross-chain swaps using hashed timelocks
• Energy efficiency of blockchain applications