To enable scalable resilient blockchain systems, several powerful general-purpose approaches toward sharding such systems have been demonstrated. Unfortunately, these approaches all come with substantial costs for ordering and execution of multi-shard transactions.

In this work, we ask whether one can achieve significant cost reductions for processing multi-shard transactions by limiting the type of workloads supported. To initiate the study of this problem, we propose core-Chimera. Core-Chimera uses strict UTXO-based environmental requirements to enable powerful multi-shard transaction processing with an absolute minimum amount of coordination between shards. In the environment we designed core-Chimera for, core-Chimera will operate perfectly with respect to all transactions proposed and approved by well-behaved clients, but does not provide any other guarantees. 

To illustrate that core-Chimera-like protocols can also be of use in environments with faulty clients, we also demonstrate two generalizations of core-Chimera, optimistic-Chimera and resilient-Chimera, that make different tradeoffs in complexity and costs when dealing with faulty behavior and attacks. Finally, we compare these three protocols and show their potential scalability and performance benefits over state-of-the-art general-purpose systems. These results underline the importance of the study of specialized approaches toward sharding in resilient systems.  


Problem: Sharded Resilient Transaction Processing with Minimal Costs

Traditional resilient systems operate on fully-replicated fault-tolerant clusters, which limits their scalability and performance. One way to make the step towards resilient high-performance systems that can deal with huge workloads is by enabling independent fault-tolerant clusters to efficiently communicate and cooperate with each other, as this also enables the usage of high-performance techniques such as sharding. Recently, such inter-cluster communication was formalized as the Byzantine cluster-sending problem. Unfortunately, existing worst-case optimal protocols for cluster-sending all have linear complexity in the size of the clusters involved. 

In this paper, we propose probabilistic cluster-sending techniques as a solution for the cluster-sending problem with only an expected constant message complexity, this independent of the size of the clusters involved and this even in the presence of highly unreliable communication. Depending on the robustness of the clusters involved, our techniques require only two-to-four message round-trips (without communication failures). Furthermore, our protocols can support worst-case linear communication between clusters. Finally, we have put our techniques to the test in an in-depth experimental evaluation that further underlines the exceptional low expected costs of our techniques in comparison with other protocols. As such, our work provides a strong foundation for the further development of resilient high-performance systems.

Solution: Byzantine Cluster-Sending in Expected Constant Cost and Constant Time

Within the Linux ecosystem, hypervisor and container-based virtualization are the two most prevalent and well-known server virtualization approaches. As it is often the case, the choice is much more complex than a binary decision between those distinct approaches. 
Recently emerging technologies, concepts and approaches, have greatly diversified the "server virtualization landscape". For example, the enabling concepts of container-based virtualization are ever changing and improve upon every upcoming Kernel release. 
Moreover, novel sandbox-based approaches leverage traditional and recent operating system functionality to intercept system calls for their isolation needs. Hybrid systems utilize classic hypervisors in order to run a specific purpose built unikernel to run container-based virtualization within themselves.

In this work, we present an approach to classify virtualization aspects by their isolation capability. For this purpose, we decompose them into their respective enabling components and describe them in detail. Finally, we present a multi-level classification of server virtualization. 
%with hints on a cross-section of them. This classification aims to enable a quick assessment of virtualization technologies and their induced implications.

SoK: Virtualization Classification on Isolation Capabilities

Some recent services use a sequencer to simplify ordering operations on sharded data. The sequencer assigns each operation a multi-sequence number which explicitly orders the operation on each shard it accesses. Existing sequencers have two shortcomings. First, failures can result in some multi-sequence numbers never being assigned, exposing a noncontiguous multi-sequence, which requires complex scaffolding to handle. Second, existing implementations use single-machine sequencers, limiting service throughput to the ordering throughput of one machine.

We make two contributions. First, we posit that sequencers should expose our new contiguous multi-sequence abstraction. Contiguity guarantees every sequence number is assigned an operation, simplifying the abstraction. Second, we design and implement MASON, the first system to expose the contiguous multi-sequence abstraction and the first to provide a scalable multi-sequence. MASON is thus an ideal building block for consistent, scalable services. Our evaluation shows MASON unlocks scalable throughput for two strongly-consistent services built on it.

JSYS 2022 Oct Papers Submissions

Problem: Sharded Resilient Transaction Processing with Minimal Costs

Solution: Byzantine Cluster-Sending in Expected Constant Cost and Constant Time

SoK: Virtualization Classification on Isolation Capabilities

Solution: Scalable, Contiguous Sequencing for Building Consistent Services