Abstract
State-of-the-art distributed in-memory datastores (FaRM, FaSST, DrTM) provide strongly-consistent distributed transactions with high performance and availability. Transactions in those systems are fully general; they can atomically manipulate any set of objects in the store, regardless of their location. To achieve this, these systems use complex distributed transactional protocols. Meanwhile, many workloads have a high degree of locality. For such workloads, distributed transactions are an overkill as most operations only access objects located on the same server – if sharded appropriately.
In this paper, we show that for these workloads, a single-node transactional protocol combined with dynamic object re-sharding and asynchronously pipelined replication can provide the same level of generality with better performance, simpler protocols, and lower developer effort. We present Zeus, an in-memory distributed datastore that provides general transactions by acquiring all objects involved in the transaction to the same server and executing a single-node transaction on them. Zeus is fault-tolerant and strongly-consistent. At the heart of Zeus is a reliable dynamic object sharding protocol that can move 250K objects per second per server, allowing Zeus to process millions of transactions per second and outperform more traditional distributed transactions on a wide range of workloads that exhibit locality.
In this paper, we show that for these workloads, a single-node transactional protocol combined with dynamic object re-sharding and asynchronously pipelined replication can provide the same level of generality with better performance, simpler protocols, and lower developer effort. We present Zeus, an in-memory distributed datastore that provides general transactions by acquiring all objects involved in the transaction to the same server and executing a single-node transaction on them. Zeus is fault-tolerant and strongly-consistent. At the heart of Zeus is a reliable dynamic object sharding protocol that can move 250K objects per second per server, allowing Zeus to process millions of transactions per second and outperform more traditional distributed transactions on a wide range of workloads that exhibit locality.
Original language | English |
---|---|
Title of host publication | Proceedings of the Sixteenth European Conference on Computer Systems (EuroSys ’21) |
Editors | Antonio Barbalace, Pramod Bhatotia, Lorenzo Alvisi, Cristian Cadar |
Publisher | ACM |
Pages | 145-161 |
Number of pages | 17 |
ISBN (Print) | 9781450383349 |
DOIs | |
Publication status | Published - 21 Apr 2021 |
Event | 16th ACM EuroSys Conference on Computer Systems - Virtual Conference, Edinburgh, United Kingdom Duration: 26 Apr 2021 → 28 Apr 2021 https://2021.eurosys.org/ |
Conference
Conference | 16th ACM EuroSys Conference on Computer Systems |
---|---|
Abbreviated title | EuroSys 2021 |
Country/Territory | United Kingdom |
City | Edinburgh |
Period | 26/04/21 → 28/04/21 |
Internet address |
Keywords / Materials (for Non-textual outputs)
- locality
- transactions
- dynamic sharding
- replication
- availability
- strict serializability
- pipelining