Blockchain technologies are gaining momentum, enabling parties to maintain global states without full trust. This paper surveys private blockchains, analyzing four dimensions: distributed ledger, cryptography, consensus protocol, and smart contract. It introduces BLOCKBENCH, a benchmarking framework to evaluate private blockchains against data processing workloads. Evaluations of Ethereum, Parity, and Hyperledger reveal performance gaps compared to databases and highlight design trade-offs. Drawing from database principles, the paper discusses research directions to improve blockchain performance. Blockchain systems are categorized into public and private. Public blockchains, like Bitcoin, use proof-of-work (PoW) for consensus, while private blockchains, such as Hyperledger, use protocols like PBFT. Public blockchains are decentralized and face scalability and security challenges, while private blockchains offer controlled environments with higher performance but require authentication. Key concepts include distributed ledgers, consensus protocols, cryptography, and smart contracts. Distributed ledgers are replicated append-only data structures. Consensus protocols vary from PoW to PBFT, with PBFT being communication-bound and efficient in private settings. Cryptography ensures ledger integrity through hash trees and cryptographic pointers. Smart contracts enable complex transaction logic, with Ethereum being Turing-complete. The paper evaluates current blockchain systems, highlighting their limitations and proposing improvements. It discusses cryptographic techniques, trusted hardware, and transaction privacy. Smart contracts are analyzed for language expressiveness and execution environments. The paper concludes that blockchain systems need to adopt database principles to enhance performance and security, with future research focusing on scalability, privacy, and efficiency.Blockchain technologies are gaining momentum, enabling parties to maintain global states without full trust. This paper surveys private blockchains, analyzing four dimensions: distributed ledger, cryptography, consensus protocol, and smart contract. It introduces BLOCKBENCH, a benchmarking framework to evaluate private blockchains against data processing workloads. Evaluations of Ethereum, Parity, and Hyperledger reveal performance gaps compared to databases and highlight design trade-offs. Drawing from database principles, the paper discusses research directions to improve blockchain performance. Blockchain systems are categorized into public and private. Public blockchains, like Bitcoin, use proof-of-work (PoW) for consensus, while private blockchains, such as Hyperledger, use protocols like PBFT. Public blockchains are decentralized and face scalability and security challenges, while private blockchains offer controlled environments with higher performance but require authentication. Key concepts include distributed ledgers, consensus protocols, cryptography, and smart contracts. Distributed ledgers are replicated append-only data structures. Consensus protocols vary from PoW to PBFT, with PBFT being communication-bound and efficient in private settings. Cryptography ensures ledger integrity through hash trees and cryptographic pointers. Smart contracts enable complex transaction logic, with Ethereum being Turing-complete. The paper evaluates current blockchain systems, highlighting their limitations and proposing improvements. It discusses cryptographic techniques, trusted hardware, and transaction privacy. Smart contracts are analyzed for language expressiveness and execution environments. The paper concludes that blockchain systems need to adopt database principles to enhance performance and security, with future research focusing on scalability, privacy, and efficiency.