A Blockchain Consensus Protocol With Horizontal Scalability

Blockchain technology has the potential to decentralise many traditionally centralised systems. However, scalability remains a key challenge. A horizontally scalable solution, where performance increases by adding more nodes, would move blockchain systems one step closer to ubiquitous use. We design a novel blockchain system called CHECO. Each node in our system maintains a personal hash chain, which only stores transactions that the node is involved in. A consensus is reached on special blocks called checkpoint blocks rather than on all transactions. Checkpoint blocks are effectively a hash pointer to the personal hash chains; thus a single checkpoint block may represent an arbitrarily large set of transactions. We introduce a validation protocol so that any node can check the validity of any transaction. Since transaction and validation protocols are point-to-point, we achieve horizontal scalability. We analytically evaluate our system and show a number of highly desirable correctness properties such as consensus on the validity of transactions. Further, we give a free and open-source implementation of CHECO and evaluate it experimentally. Our results show a strong indication of horizontal scalability.

Kelong Cong

Blockchain systems have the potential to decentralise many traditionally centralised systems. However, scalability remains a key challenge. Without a horizontally scalable solution, where performance increases by adding more nodes to the system, blockchain systems remain unsuitable for ubiquitous use. We design a novel blockchain system called Checo. Each node in our system maintains a personal hash chain, which only stores transactions that the node is involved in. A consensus is reached on special blocks called checkpoint blocks rather than on all the transactions. Checkpoint blocks are effectively a hash pointer to the personal hash chains; thus a single checkpoint block may represent an arbitrarily large set of transactions. We introduce a validation protocol so that any node can check the validity of any transaction. Since it is a point-to-point protocol, we achieve horizontal scalability. We analytically evaluate our system and proof a number of highly desirable correctness properties. Further, we give a free and open-source implementation of Checo and evaluate it experimentally. Our results show a strong indication of horizontal scalability.

2017

Manycore Network Interfaces for In-Memory Rack-Scale Computing

Edouard Bugnion, Alexandros Daglis, Babak Falsafi, Boris Robert Grot, Stanko Novakovic

Datacenter operators rely on low-cost, high-density technologies to maximize throughput for data-intensive services with tight tail latencies. In-memory rack-scale computing is emerging as a promising paradigm in scale-out datacenters capitalizing on commodity SoCs, low-latency and high-bandwidth communication fabrics and a remote memory access model to enable aggregation of a rack’s memory for critical data-intensive applications such as graph processing or key-value stores. Low latency and high bandwidth not only dictate eliminating communication bottlenecks in the software protocols and off-chip fabrics but also a careful on-chip integration of network interfaces. The latter is a key challenge especially in architectures with RDMA-inspired one-sided operations that aim to achieve low latency and high bandwidth through on-chip Network Interface (NI) support. This paper proposes and evaluates network interface architectures for tiled manycore SoCs for in-memory rack-scale computing. Our results indicate that a careful splitting of NI functionality per chip tile and at the chip’s edge along a NOC dimension enables a rack-scale architecture to optimize for both latency and bandwidth. Our best manycore NI architecture achieves latencies within 3% of an idealized hardware NUMA and efficiently uses the full bisection bandwidth of the NOC, without changing the on-chip coherence protocol or the core’s microarchitecture.

2015

Secure, Confidential Blockchains Providing High Throughput and Low Latency

Eleftherios Kokoris Kogias

One of the core promises of blockchain technology is that of enabling trustworthy data dissemination in a trustless environment. What current blockchain systems deliver, however, is slow dissemination of public data, rendering blockchain technology unusable in settings where latency, transaction capacity, or data confidentiality are important. In this thesis we focus on providing solutions on two of the most pressing problems blockchain technology currently faces: scalability and data confidentiality. To address the scalability issue, we present OMNILEDGER, a novel scale-out distributed ledger that preserves long-term security under permissionless operation. It ensures security and correctness by using a bias-resistant public-randomness protocol for choosing large, statistically representative shards that process transactions, and by introducing an efficient cross-shard commit protocol that atomically handles transactions affecting multiple shards. To enable secure sharing of confidential data we present CALYPSO, the first fully decentralized, auditable access-control framework for secure blockchain-based data sharing which builds upon two abstractions. First, on-chain secrets enable collective management of (verifiably shared) secrets under a Byzantine adversary where an access-control blockchain enforces user-specific access rules and a secret-management cothority administers encrypted data. Second, skipchain-based identity and access management enables efficient administration of dynamic, sovereign identities and access policies and, in particular, permits clients to maintain long-term relationships with respect to evolving user identities thanks to the trust-delegating forward links of skipchains. In order to build OMNILEDGER and CALYPSO, we first build a set of tools for efficient decentralization, which are presented in Part II of this dissertation. These tools can be used in decentralized and distributed systems to achieve (1) scalable consensus (BYZCOIN), (2) bias- resistant distributed randomness creations (RANDHOUND), and (3) relationship-keeping between independently updating communication endpoints (SKIPCHAINIAC). Although we use this tools in the scope off this thesis, they can be (and already have been) used in a far wider scope.

EPFL2019

Secure, Confidential Blockchains Providing High Throughput and Low Latency

Eleftherios Kokoris Kogias

EPFL2019