Theory & Proposal

You've probably heard about the recent uproar over Solana, a permissionless blockchain, struggling with network congestion from bot meme coin transactions. ^{In due time I’m confident they’ll fix that issue but that’s another conversation.} Well, this group of bottleneck detectives are tackling a similar - but distinct - issue with permissioned blockchains like Hyperledger Fabric and Quorum.

Rather than dealing with spam attacks, they're on the hunt for performance bottlenecks - they’re holding back a blockchain's throughput from reaching its full potential. The theory? To develop an approach to sniff out bottlenecks by systematically reviewing past work, running experiments on the Distributed Ledger Performance Scan (DLPS), and cooking up fancy visualizations. A game of blockchain forensics.

^{Had to take the time to insert an inspector gadget reference.}

Findings

Here is an outline of their method:

1. Running experiments using the Distributed Ledger Performance Scan (DLPS), which is an open-source blockchain benchmarking framework.

2. Extending and utilizing the graphical evaluation capabilities of the DLPS framework.

3. Applying exploratory data analysis techniques to the resource utilization metrics and performance data collected from the DLPS experiments.

4. Visually analyzing trends and correlations in the graphed data to pinpoint potential bottleneck components within the blockchain systems.

After putting this bottleneck-detecting method through its paces on Hyperledger Fabric and Quorum, the came up with some juicy findings to report.