Guest lectures Monday March 21st and Wednesday March 23rd, 2022: Professor Lorenzo Alvisi, Cornell University

We are excited to announce that next week we will have a visit from Professor Lorenzo Alvisi of Cornell. He will be giving guest lectures on Monday March 21st and Wednesday March 23rd, 2022. These lectures will be held in person. They will also be streamed on Zoom simultaneously for those who cannot attend physically but physical presence will be appreciated.

Lorenzo Alvisi on an Italien Ducati.
Lorenzo Alvisi on an Italien Ducati. Foto: Privat.
Portrettbilde av Fuglesteg, Jan
Fuglesteg, Jan Studierådgiver
Publisert: 15.03.22 13:31 Oppdatert: 15.03.22 14:02
Naturvitenskap Studentliv / Studier Teknologi

Professor Alvisi is an outstanding researcher who focuses on distributed systems and dependability, especially Byzantine fault tolerance. He has co-authored many award-winning papers in top-tier computer science conferences, including Best Paper awards at OSDI 2020SYSTOR 2014SOSP 2007USENIX 2007and more.

The Monday talk is titled Orderrrrr!!! A tale of money, intrigue, and specifications.
The Wednesday talk is titled Towards a high performance shared log.

See below for more info.

Professional contact person is Professor Otto Anshus.

All interested are most welcome!

Date: Monday March 21st, 2022, 14:15
Physical lecture: Teknologibygget 1.023 (map).
Digital: Zoom-link. Meeting ID: 682 9276 8239. Password: 099180.
Title: "Orderrrrr!!! A tale of money, intrigue, and specifications"
Lecturer: Professor Lorenzo Alvisi, Cornell University

Mistrust over traditional financial institutions is motivating the development of decentralized financial infrastructures based on blockchains. In particular, Consortium blockchains (such as the Linux Foundation Hyperledger and Facebook’s diem) are emerging as the approach preferred by businesses. These systems allow only a well-known set of mutually distrustful parties to add blocks to the blockchain; in this way, they aim to retain the benefits of decentralization without embracing the cyberpunk philosophy that informed Nakamoto’s disruptive vision. At the core of consortium blockchains is State Machine Replication, a classic technique borrowed from fault tolerant distributed computing; to ensure the robustness of their infrastructure, consortium blockchains actually borrow the Byzantine-tolerant version of this technique, which guarantees that the blockchain will operate correctly even if as many as about a third of the contributing parties are bent on cheating.

But, sometimes, “a borrowing is a sorrowing”.

I will discuss why Byzantine-tolerant state machine replication is fundamentally incapable of recognizing, never mind preventing, an ever present scourge of financial exchanges: the fraudulent manipulation of the order in which transactions are processed — and how its specification needs to be expanded to give it a fighting chance.

But is it possible to completely eliminate the ability of Byzantine parties to engage in order manipulation? What meaningful ordering guarantees can be enforced? And at what cost?

Date: Wednesday March 23rd, 2022, 14:15
Physical lecture: TMH Auditorium 2 (U6.A2C) (map).
Digital: Zoom-link. Meeting ID: 682 9276 8239. Password: 099180.
Title: "Towards a high performance shared log"
Lecturer: Professor Lorenzo Alvisi, Cornell University

The shared ledger abstraction used by blockchains to support decentralized finance applications, supply-chain management, fraud detection, and more, is more broadly emerging as a key building block for distributed systems. Shared ledgers (often referred to as logs) underlie, among others, state machine replication, atomic transactions, stream processing, intrusion detection, and failure recovery, while the abstraction’s simplicity and generality make it popular for developing reliable cloud applications based on microservices or serverless computing.

An ideal shared log would guarantee high throughput (through sharding) and low latency, remain available during reconfiguration (when shards are removed or added in response to failures or to the changing needs of its clients) and offer strong consistency guarantees, such as total order and linearizability [Her1990]. Unfortunately, no existing shared log matches this ideal, and existing shared log implementations are forced to make painful tradeoffs. We have made some progress towards this ideal — and I’ll share with you where we are at in this work in progress.

About Lorenzo Alvisi
Lorenzo Alvisi is a Tisch University Professor at Cornell University. Prior to joining Cornell, he was a University Distinguished Teaching Professor and the holder of the Endowed Professorship #5 at University of Texas at Austin. His research focuses on distributed systems and dependability. He holds a baccalaureate in Physics from the University of Bologna (1987), and an MS and PhD in Computer Science from Cornell University (1994 and 1996, respectively). He is a Fellow of the Institute of Electrical and Electronics Engineers and of the Association for Computing Machinery.

He is a co-author of many award-winning papers in top computer science conferences, including Best Paper awards at OSDI 2020SYSTOR 2014SOSP 2007USENIX 2007and more.

Kortnytt fra Institutt for informatikk
Fuglesteg, Jan Studierådgiver
Vi anbefaler