Antonios Katsarakis
Research Subjects
Research Interests
Antonios Katsarakis has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Angelos Bilas
Institute of Computer Science, Foundation for Research and Technology - Hellas (FORTH), Heraklion, Crete, Greece and Computer Science Department, University of Crete, Heraklion, Crete, Greece
1 shared publication - Atsushi Koshiba
Technical University of Munich, Munich, Germany
1 shared publication - Chris Jensen
Cambridge University, Cambridge, UK and Microsoft Research Cambridge
1 shared publication - Dimitra Giantsidi
Azure Research, Cambridge, United Kingdom
1 shared publication - Felix Gust
Technical University of Munich, Munich, Germany
1 shared publication - Giorgos Saloustros
FORTH, Heraklion, Crete, Greece
1 shared publication
Latest Publications
- Open Access
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 43-49
Compaction Size and Tail Latency in LSM-based Key-Value Stores
- 26 April 2026
- Published by Association for Computing Machinery (ACM)
- p. 43-49
Open AccessCompaction Size and Tail Latency in LSM-based Key-Value Stores
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1282-1301
TNIC: A Trusted NIC Architecture: A hardware-network substrate for building high-performance trustworthy distributed systems
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1282-1301
Open AccessTNIC: A Trusted NIC Architecture: A hardware-network substrate for building high-performance trustworthy distributed systems
- Open Access
- 3 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 186-199
DLHT: A Non-blocking Resizable Hashtable with Fast Deletes and Memory-awareness
This paper presents DLHT, a concurrent in-memory hashtable. Despite efforts to optimize hashtables, that go as far as sacrificing core functionality, state-of-the-art designs still incur multiple memory accesses per request and block request processing in three cases. First, most hashtables block while waiting for data to be retrieved from memory. Second, open-addressing designs, which represent the current state-of-the-art, either cannot free index slots on deletes or must block all requests to do so. Third, index resizes block every request until all objects are copied to the new index. Defying folklore wisdom, DLHT forgoes open-addressing and adopts a fully-featured and memory-aware closed-addressing design based on bounded cache-line-chaining. This design offers (1) lock-free operations and deletes that free slots instantly, (2) completes most requests with a single memory access, (3) utilizes software prefetching to hide memory latencies, and (4) employs a novel non-blocking and parallel resizing. In a commodity server and a memory-resident workload, DLHT surpasses 1.6B requests per second and provides 3.5× (12×) the throughput of the state-of-the-art closed-addressing (open-addressing) resizable hashtable on Gets (Deletes).
- 3 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 186-199
Open AccessDLHT: A Non-blocking Resizable Hashtable with Fast Deletes and Memory-awareness
This paper presents DLHT, a concurrent in-memory hashtable. Despite efforts to optimize hashtables, that go as far as sacrificing core functionality, state-of-the-art designs still incur multiple memory accesses per request and block request processing in three cases. First, most hashtables block while waiting for data to be retrieved from memory. Second, open-addressing designs, which represent the current state-of-the-art, either cannot free index slots on deletes or must block all requests to do so. Third, index resizes block every request until all objects are copied to the new index. Defying folklore wisdom, DLHT forgoes open-addressing and adopts a fully-featured and memory-aware closed-addressing design based on bounded cache-line-chaining. This design offers (1) lock-free operations and deletes that free slots instantly, (2) completes most requests with a single memory access, (3) utilizes software prefetching to hide memory latencies, and (4) employs a novel non-blocking and parallel resizing. In a commodity server and a memory-resident workload, DLHT surpasses 1.6B requests per second and provides 3.5× (12×) the throughput of the state-of-the-art closed-addressing (open-addressing) resizable hashtable on Gets (Deletes).
- Open Access
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 44-57
Unanimous 2PC: Fault-tolerant Distributed Transactions Can be Fast and Simple
Distributed transactional datastores are pivotal in supporting the needs of modern applications and services. Datastores rely on distributed transactional protocols to tolerate faults while also aiming for strong consistency and high performance. State-of-the-art transactional protocols, such as FaRM, provide a lock-free execution phase, thus improving performance and simplifying recovery. However, these protocols often come with lengthy and complicated commit phases, requiring multiple round-trips to commit a transaction (or additional replicas). This completely contrasts with the simplicity and efficiency of the traditional two-phase commit (2PC) protocol, which can commit a transaction after a single round-trip, albeit lacking fault tolerance. To address the limitations of both approaches, we introduce U2PC, a novel 2PC variant that combines simplicity, efficiency, and fault tolerance. U2PC frugally replicates data (f + 1) times to tolerate up to f faults with strong consistency. It offers a single round-trip commit after unanimous responses of all replicas of the involved shards and ensures safe recovery via an extra "pre-abort" round before aborting a transaction. Our verification in TLA+ confirms that U2PC achieves strict serializability and recovers safely. In short, U2PC ensures fault tolerance, optimizes performance for common scenarios, and offers uniform transaction handling.
- 22 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 44-57
Open AccessUnanimous 2PC: Fault-tolerant Distributed Transactions Can be Fast and Simple
Distributed transactional datastores are pivotal in supporting the needs of modern applications and services. Datastores rely on distributed transactional protocols to tolerate faults while also aiming for strong consistency and high performance. State-of-the-art transactional protocols, such as FaRM, provide a lock-free execution phase, thus improving performance and simplifying recovery. However, these protocols often come with lengthy and complicated commit phases, requiring multiple round-trips to commit a transaction (or additional replicas). This completely contrasts with the simplicity and efficiency of the traditional two-phase commit (2PC) protocol, which can commit a transaction after a single round-trip, albeit lacking fault tolerance. To address the limitations of both approaches, we introduce U2PC, a novel 2PC variant that combines simplicity, efficiency, and fault tolerance. U2PC frugally replicates data (f + 1) times to tolerate up to f faults with strong consistency. It offers a single round-trip commit after unanimous responses of all replicas of the involved shards and ensures safe recovery via an extra "pre-abort" round before aborting a transaction. Our verification in TLA+ confirms that U2PC achieves strict serializability and recovers safely. In short, U2PC ensures fault tolerance, optimizes performance for common scenarios, and offers uniform transaction handling.