On Scalability and Performance of Permissioned Blockchain

Systems
Chrysoula Stathakopoulou
IBM Research - Zürich
Departement of Computer Science - ETH Zürich
[email protected]
ABSTRACT Castro and Liskov [6] allow a large throughput. Their high message
Blockchain systems have emerged as the backbone of the Bitcoin complexity, though, limits their scalability. As Vukolić analyzes [16],
cryptocurrency. Even though considerable research has been made numerous protocols have been proposed as improvements to both
towards Proof of Work based blockchain systems, the scalability PoW and BFT and there exist efforts to combine the best of both
of systems based on Byzantine Fault Tolerance (BFT) protocols is worlds.
still limited. This work aims to suggest BFT consensus protocols
for scalable blockchain systems that allow reconfiguration. At the 3 RESEARCH PLAN
same time the participants of the system are considered rational and So far, considerable research has been conducted in the perfor-
incentive mechanisms are adopted in the blockchain ecosystem. mance and security of permissionless blockchain systems. Decker
and Wattenhofer [7] study the time needed for the propagation
1 INTRODUCTION of a transaction block in the Bitcoin network and they associate
Blockchain systems have emerged as the backbone of the Bitcoin the delay with the size of the block. Gervais et al. [9] developed a
cryptocurrency, introduced in 2008 [14]. A blockchain system con- framework to evaluate and compare Bitcoin with Litecoin, Doge-
sists of a distributed append-only ledger of blocks of transactions. coin, Bitcoin’s most important forks, as well as with Ethereum.
The participants of the cryptocurrency system need to maintain a However, the recently conceptualized permissioned blockchain
synchronized view of the ledger and, therefore, a consensus proto- systems have not been studied until now in depth. Therefore, the
col is required. Bitcoin introduced a Proof of Work (PoW) consensus proposed research project aims to explore this area. As a first step,
mechanism where, essentially, each participant votes with its com- existing consensus protocols must be deployed in the context of a
puting power. blockchain system and in the scale of a wide area network so that
In 2014 Ethereum Project [17], a new cryptocurrency, redefined network parameters such as latency, throughput, availability and
blockchain systems introducing smart contracts, which are custom scalability can be evaluated. Taking into account the results of this
application code that enforces properties among the blockchain study, the goal is to suggest an optimal protocol according to a set
participants. A cryptocurrency exchange is a special case of a smart of requirements (security assumptions, number of nodes, network
contract. The concept of smart contracts decoupled the currency topology, transaction and block size, etc.) that correspond to a
logic from the distributed ledger infrastructure and opened the realistic enterprise blockchain ecosystem. Looking into enterprise
way to general purpose blockchain systems. This resulted in the blockchain proof-of-concepts and prototypes, already reveals a
conception of private or enterprise blockchain systems, such as wide variety of requirements and therefore the suggested protocol
Hyperledger Fabric [5], Tendermint [3], Chain [1], and Quorum [2]. should be highly configurable.
Unlike Ethereum, Bitcoin and other cryptocurrencies, where any- More specifically, regarding the scalability parameter, despite
one can participate, enterprise blockchain systems enforce access the extended aforementioned research in Byzantine Fault Toler-
control to the participants and, hence, we call such a blockchain ance, defining and evaluating a practical protocol that scales up to
system permissioned, whereas public systems like Bitcoin are called thousand nodes, a realistic number of stakeholders in an enterprise
permissionless. blockchain setup, is still an open problem. This research work aims
to explore this problem.
2 MOTIVATION Considering the configuration of the system, in a realistic sce-
Enterprise blockchain systems have different properties and re- nario the set of participants should be able to change without com-
quirements. Proof of Work mechanisms are not sustainable as they promising security. However, unlike PoW consensus BFT protocols
waste exorbitant amounts of energy. More recent protocols [10] [11] require a static set of known participants. To this direction, Mar-
move to Proof of Stake consensus mechanisms to tackle the energy tin and Alvisi [13] and later Rodrigues et al. [15] have suggested
inefficiency. However, both mechanisms suffer from low through- automatic reconfiguration protocols for reliable distributed stor-
put and high latency, as a result of the probabilistic agreement, age systems. This work aims to further explore dynamic consen-
which makes them unsuitable for an enterprise ecosystem. On the sus protocols and evaluate them in the context of a permissioned
other hand, while open membership systems are prone to Sybil blockchain system. Hyperledger Fabric [5] has already decoupled
attacks [8], the access control in permissioned systems allows us to ordering from execution and validation of transactions. As a next
move towards Byzantine Fault Tolerant (BFT) algorithms [12] that step, we envision a public yet permissioned distributed ordering
have been studied for decades now with respect to the machine service, where participants can offer resources for the ordering
state replication problem. Practical BFT protocols, introduced by and receive payment for their services in the form of transaction
EuroSys Doctoral Workshop ’18, April 23, 2018, Porto, Portugal C. Stathakopoulou

fees. The dynamic reconfigurable protocol is required to serve this [13] J.-P. Martin and L. Alvisi. A framework for dynamic byzantine storage. In
scenario. Dependable Systems and Networks, 2004 International Conference on, pages 325–
334. IEEE, 2004.
Finally, in a blockchain ecosystem the participants of the consen- [14] S. Nakamoto. Bitcoin: A peer-to-peer electronic cash system. https://bitcoin.org/
sus protocol can belong to different organisations with opposing bitcoin.pdf, 2008.
[15] R. Rodrigues, B. Liskov, K. Chen, M. Liskov, and D. Schultz. Automatic reconfigu-
interests, unlike in a traditional distributed system. In such an en- ration for large-scale reliable storage systems. IEEE Transactions on Dependable
vironment, each participant is expected to be rational, i.e. try to and Secure Computing, 9(2):145–158, 2012.
optimize its own state. Therefore, the participants must be given [16] M. Vukolić. The quest for scalable blockchain fabric: Proof-of-work vs. bft
replication. In International Workshop on Open Problems in Network Security,
incentives to comply with the protocol. Moreover, if we consider a pages 112–125. Springer, 2015.
public ordering service, the participants should be incentivised to [17] G. Wood. Ethereum: A secure decentralised generalised transaction ledger.
actively participate in the protocol, despite their negative utility i.e. http://gavwood.com/paper.pdf, 2014.

the cost of the resources they dedicate. Solidus [4] is a recent pro-
posal for incentive-compatible Proof of Work consensus in a public
blockchain, while Ouroboros [11] studies rational participants with
respect to a Proof of Stake Protocol. This research project aims to
study how to adopt Game Theory mechanisms in a BFT protocol.

4 SUMMARY
To summarize, this research work will focus on the performance and
scalability of distributed fault tolerant protocols with application
to blockchain systems, targeting the following key points:
(1) Analysis and large scale evaluation of the performance, scal-
ability and security of consensus mechanisms. Suggestion
of optimal protocols with respect to required network and
security parameters.
(2) Suggestion and evaluation of dynamic consensus mecha-
nisms that allow the reconfiguration of the network without
compromising the security properties of the system, while
allowing a public ordering service.
(3) Incentive compatible BFT protocols.

REFERENCES
[1] Chain. http://chain.com.
[2] Quorum. http://www.jpmorgan.com/global/Quorum.
[3] Tendermint. http://tendermint.com.
[4] I. Abraham, D. Malkhi, K. Nayak, L. Ren, and A. Spiegelman. Solidus: An incentive-
compatible cryptocurrency based on permissionless byzantine consensus. arXiv
preprint arXiv:1612.02916, 2016.
[5] E. Androulaki, A. Barger, V. Bortnikov, C. Cachin, K. Christidis, A. De Caro,
D. Enyeart, C. Ferris, G. Laventman, Y. Manevich, S. Muralidharan, C. Murthy,
B. Nguyen, M. Sethi, G. Singh, K. Smith, A. Sorniotti, C. Stathakopoulou,
M. Vukolić, S. W. Cocco, and J. Yellick. Hyperledger fabric: A distributed operat-
ing system for permissioned blockchains. In EuroSys 2018: Thirteenth EuroSys
Conference. ACM, 2018.
[6] M. Castro, B. Liskov, et al. Practical Byzantine fault tolerance. In Proc. Third
USENIX Symposium on Operating Systems Design and Implementation (OSDI), New
Orleans, Louisiana, USA, February 22-25, 1999, pages 173–186. USENIX Association,
1999.
[7] C. Decker and R. Wattenhofer. Information propagation in the bitcoin network.
In Peer-to-Peer Computing (P2P), 2013 IEEE Thirteenth International Conference on,
pages 1–10. IEEE, 2013.
[8] J. R. Douceur. The sybil attack. In International workshop on peer-to-peer systems,
pages 251–260. Springer, 2002.
[9] A. Gervais, G. O. Karame, K. Wüst, V. Glykantzis, H. Ritzdorf, and S. Capkun. On
the security and performance of proof of work blockchains. In Proceedings of the
2016 ACM SIGSAC Conference on Computer and Communications Security, pages
3–16. ACM, 2016.
[10] Y. Gilad, R. Hemo, S. Micali, G. Vlachos, and N. Zeldovich. Algorand: Scaling
byzantine agreements for cryptocurrencies. In Proceedings of the 26th Symposium
on Operating Systems Principles, pages 51–68. ACM, 2017.
[11] A. Kiayias, A. Russell, B. David, and R. Oliynykov. Ouroboros: A provably
secure proof-of-stake blockchain protocol. In Annual International Cryptology
Conference, pages 357–388. Springer, 2017.
[12] L. Lamport, R. Shostak, and M. Pease. The byzantine generals problem. ACM
Transactions on Programming Languages and Systems (TOPLAS), 4(3):382–401,
1982.