Javad Saberlatibari

dblp:321/5805 · DBLP profile ↗
← Back
2ranked-venue papers
0as first author
2since 2021 · last 2026
0000-0002-3968-867XORCID · corroborated

Domains — the database's venue-derived domains; a paper can count in several

Software engineering, systems software and programming languages · 2 · 2 since 2021

Expertise — from the expertise taxonomy: the topics of the expert's papers under the CCF categories. A weight counts papers with recency: 1 for a paper about the topic, 0.3 when the topic is its context, halved every five years.

Computer architecture, parallel and distributed computing, and storage systems
1 paper
Distributed systems · 56% Memory systems · 28% Cloud and datacenter computing · 8%

Topics — the 5 heaviest of 5, each with the papers that count most for it

TopicWeightPapersLastEvidence papers
Memory systems › memory consistency › memory consistency model
hybrid consistency
0.612022
Hamband: RDMA replicated data types · PLDI 2022
Distributed systems › replication
replicated data types
0.612022
Hamband: RDMA replicated data types · PLDI 2022
Distributed systems
replication
0.612022
Hamband: RDMA replicated data types · PLDI 2022
Cloud and datacenter computing
datacenter network
0.212022
Hamband: RDMA replicated data types · PLDI 2022
Interconnection networks and networks-on-chip
remote direct memory access
0.212022
Hamband: RDMA replicated data types · PLDI 2022

Methods — techniques the papers use, named apart from their topics

operational semantics · 0.6formal verification · 0.6
YearPublicationVenuePosition
2026 Frashokereti: Non-aborting Optimistically Replicated Objects
abstract
Optimistic replication of objects avoids coordination and brings higher responsiveness and availability. However, when clients issue concurrent operations, conflicts naturally arise which can lead the replicated states to diverge or lose integrity. When conflicts occur, existing approaches resort to pessimism or abortion. This paper characterizes ORDTs (Optimistically Replicated Data Types), objects that can be optimistically replicated with convergence and integrity, and without aborting calls. It shows that ORDTs subsume CRDTs and transformed relational schema, and presents techniques to convert objects to ORDTs. It further proves that optimistic replication for objects that fall out of ORDTs is aborting and NP-Complete. Further, it presents an optimistic replication protocol for ORDTs called Frashokereti . It uses a statically decided order to efficiently order calls. The paper proves that Frashokereti is sound for every ORDT, i.e., Frashokereti is optimistic and non-aborting, and preserves convergence, integrity, and liveness properties. Experimental results show that Frashokereti significantly outperforms previous optimistic protocols.
Eric Man Chan, Javad Saberlatibari, Mohsen Lesani
Proc. ACM Program. Lang.2
2022 Hamband: RDMA replicated data types
abstract
Data centers are increasingly equipped with RDMAs. These network interfaces mark the advent of a new distributed system model where a node can directly access the remote memory of another. They have enabled microsecond-scale replicated services. The underlying replication protocols of these systems execute all operations under strong consistency. However, strong consistency can hinder response time and availability, and recent replication models have turned to a hybrid of strong and relaxed consistency. This paper presents RDMA well-coordinated replicated data types, the first hybrid replicated data types for the RDMA network model. It presents a novel operational semantics for these data types that considers three distinct categories of methods and captures their required coordination, and formally proves that they preserve convergence and integrity. It implements these semantics in a system called Hamband that leverages direct remote accesses to efficiently implement the required coordination protocols. The empirical evaluation shows that Hamband outperforms the throughput of existing message-based and strongly consistent implementations by more than 17x and 2.7x respectively.
Farzin Houshmand, Javad Saberlatibari, Mohsen Lesani
PLDI2