We covered DLT and blockchain in Chapter 1, Exploring Blockchain and BaaS, and Chapter 2, Construing Distributed Ledger Tech and Blockchain. In addition, we learned about various network topologies, such as centralized, distributed, and decentralized systems. We also became familiar with the structure of blockchain, transactions, and various other blockchain concepts. We carried out a detailed analysis of, and discussion on, permissioned and permissionless blockchain. Permissionless blockchain, such as Ethereum and Bitcoin, are open blockchain, where anyone can participate. On the other hand, permissioned blockchain allows a limited set of participants to administer the blockchain network, while only authorized and authenticated sets of participants are able to access it.

There are various advantages associated with permissionless blockchain, and similarly, there are advantages to using permissioned blockchain. Permissioned blockchain is cost-effective, and has low transaction overheads. As transaction verification and validation is faster, transaction costs are very low and transaction times are faster. The decision behind the choice of blockchain network depends entirely on the use case and the visibility of messages and transactions. However, in my opinion, the key difference that resonates well with enterprises is the determination as to who will participate in the blockchain business network and who is authorized to transact on the business network; another reason being the ability to empower the direct relationship between producer and consumer, and reducing or removing the reliance on middle parties or third parties (intermediaries).

We are in the era of disintermediation. It has been pioneered by Uber, Amazon, Airbnb, and others, where they own no vehicles, hold no real inventory, and have no inventory of rooms, respectively, yet they allow producers and consumers to connect and transact. Blockchain and DLT further empower producers and consumers, resulting in the disintermediation of third parties and intermediaries from the equation. Permissioned blockchain enhances privacy by means of permissioned access to blockchain networks and channeling transactions between participants to further allow the segregation of data on blockchain networks, thereby enabling privacy and the confidentiality of data. DLT, like blockchain, has the full potential to disrupt industries that have a high reliance on intermediaries, such as insurance, healthcare, transportation, retail, logistics, real estate, and education.

At a high level, with regard to an enterprise, blockchain offers the following:

• It checks for the risk of malicious nodes tampering with data in the event of transmission to ensure tamper-proof data transmission. This is ensured by securing the transaction tree, along with complexity in gaining PoW. Malicious users cannot alter/tamper with data without recomputing the PoW hash, which is a gigantic task and requires extreme computing power.
• As the HLF blockchain is permissioned, the blockchain network is operating among known participants, which offers a high level of trust in the blockchain network itself.
• Using HLF's channels, transactions between groups of participants can be secured further.
• It removes the reliance on a single, central point of failure.
• Consistency is ensured by following protocol, adopting the same rules for validation and block layout.
• All nodes will follow the longest chain, which ensures the establishment of agreement across geographies.
• Blockchain lowers uncertainty and enhances trust between parties, leading to faster and more secure transactions.
• Permissioned blockchain, chosen by enterprises, allows enterprises to define membership rules for participants by providing immutability (tamper-proof, where blockchain represents the truth), privacy, and confidentiality (the secure exchange of sensitive data with authorization), scalability, reliability, availability (to support mission-critical applications), and auditability (manage, track, trace, verify, and monitor).

In a permissionless blockchain, transactions are executed on every node (assuming the consensus is PoW). This means a lack of confidentiality because data and smart contracts, as well as transaction data, is available on every node on the network. Confidentiality and transaction data visibility is of great importance for enterprises. In the case of B2B transactions, an enterprise would not like the data on special rates that is offered to one partner being available to another partner, although confidentiality is addressed in permissionless networks by encrypting data. However, permissionless blockchain networks using PoW will lead to data being available on every node, which highlights the possibility of decrypting it, given time, and the local availability of data on the node. In HLF, along with the participants identified and encryption, channels offer the highest level of confidentiality. Here, only participating nodes will have access to chaincode and transaction data and that, too, is further controlled by access control. This introduces a high level of privacy and confidentiality to a blockchain network.

Permissionless versus permissioned blockchain:

When we enter into a discussion on Why Hyperledger?, it makes sense to quickly look into the differences between permissionless blockchain and permissioned blockchain, such as HLF. Let's analyze these variants of DLT in terms of execution style, determinism, and confidentiality:

• Execution style:
  • A permissionless blockchain, such as Ethereum, observes the sequential execution of transactions, where they abide by the order-execute architecture. All peers execute the order-execute style of transactions and this results in performance and scalability limitations. Here, the throughput is inversely proportional to latency in transaction. However, permissionless blockchain tries to handle this by orchestrating around a cryptocurrency. This ensures that a fuel/gas is included with each transaction. Hence, a gas is paid for each step of the transaction execution via a smart contract. However, such a mechanism of engulfing a cryptocurrency might not fit into the permissioned blockchain.
  • A permissioned blockchain, like HLF's architecture, supports scalability and performance, along with trust. HLF's architecture is based on execute-order-validate (E-O-V) architecture, where transactions are executed even before a consensus is reached. Execution (execute) of the transaction will ensure a transaction's correctness (endorsement), while a modular pluggable consensus protocol will result in an ordering (order) transaction. Furthermore, just before committing the transaction, it is validated (validate) by an application-specific endorsement policy. E-O-V addresses the flexibility, scalability, performance, and confidentiality issues faced by the order and execute architecture of permissionless blockchain. HLF allows a subset of peers to execute transactions in parallel. Interestingly, chaincode delegates the work of endorsement to certain designated peers; hence, different chaincode can designate different peers as endorsers, which supports parallel execution. Note that Fabric executes a transaction even before it is ordered.
• Determinism:
  • Consensus ensures that nodes are in agreement over a transaction and, hence, the smart contract should execute transactions deterministically. If not, there is no point in establishing a consensus. In addition, such non-determinism will lead to nullifying the consensus, and this will result in forks. Hence, smart contract languages and compilers should ensure that smart contract execution is deterministic. Hence, various blockchains opt for DSL. This forces developers to learn new languages, just to ensure the determinism of smart contracts.
  • HLF's E-O-V architecture ensures that a transaction is validated by an application-specific endorsement policy. This means that it is the application-specific policy that ensures how many, and which, peer nodes will validate and ensure the deterministic execution of the chaincode. Hence, a subset of peers will execute (endorse) the transaction to meet the endorsement policy. This will filter out inconsistent results, even before ordering, and thereby eliminate any non-determinism. Because non-determinism is eliminated, HLF supports the use of a standard programming language. You can write chaincode in Go, Node.js, and Java.
• Hybrid replication driver for determinism:
  • HLF follows passive and active replication. Passive replication is achieved by executing (endorsing) a transaction by a subset of peers, which offers determinism and parallel execution. It also achieves active replication by committing transactions to ledger only after a consensus is reached. Hence, HLF follows a hybrid replication strategy. Again, the choice of consensus is specific to the use case, or deployment in relation to that use case, as Hyperledger supports a modular consensus mechanism. This allows implementers to choose any protocol of choice for consensus, such as BFT or CFT.
• Confidentiality:
  • Permissionless blockchain, which leverages PoW, executes transactions on every node. Hence, every transaction and the smart contract are visible to each node, which clearly indicates a loss of confidentiality for the gain of BFT offered by PoW. Loss of confidentiality is a challenge for enterprise customers and their use cases. For example, if a business wanted to establish certain rates with some suppliers and a different rate with non-premium suppliers, they would not be able to maintain the confidentiality of such preferred rates. If all suppliers are on the same blockchain network and access the same smart contract, it is impossible to maintain different trade relationships (rates) with different suppliers. Permissionless blockchain offers two types of solution:
    • It encrypts such preferred information. However, data and smart contracts are on every node. Encryption can be compromised and there is always the risk of losing information.
    • Zero knowledge proofs (ZKP) can handle a loss of confidentiality. However, ZKP's computation increases latency and consumes resources. This means that ZKP can solve confidentiality issues, even though this will lead to performance issues.
• A permissioned blockchain, like HLF, offers channels and private data collection (PDC) to address the confidentiality issue.