Abstract | ||
---|---|---|
There is a trend in recent database research to pursue coordination avoidance and weaker transaction isolation under a long-standing assumption: concurrent serializable transactions under read-write or write-write conflicts require costly synchronization, and thus may incur a steep price in terms of performance. In particular, distributed transactions, which access multiple data items atomically, are considered inherently costly. They require concurrency control for transaction isolation since both read-write and write-write conflicts are possible, and they rely on distributed commitment protocols to ensure atomicity in the presence of failures. This paper presents serializable read-only and write-only distributed transactions as a counterexample to show that concurrent transactions can be processed in parallel with low-overhead despite conflicts.
Inspired by the slotted ALOHA network protocol, we propose a simpler and leaner protocol for serializable read-only write-only transactions, which uses only one round trip to commit a transaction in the absence of failures irrespective of contention. Our design is centered around an epoch-based concurrency control (ECC) mechanism that minimizes synchronization conflicts and uses a small number of additional messages whose cost is amortized across many transactions. We integrate this protocol into ALOHA-KV, a scalable distributed key-value store for read-only write-only transactions, and demonstrate that the system can process close to 15 million read/write operations per second per server when each transaction batches together thousands of such operations.
|
Year | DOI | Venue |
---|---|---|
2017 | 10.1145/3127479.3127487 | SoCC '17: ACM Symposium on Cloud Computing
Santa Clara
California
September, 2017 |
Keywords | Field | DocType |
Epoch-based concurrency control, consistency, distributed transaction, multi-put or multi-get key-value store | Compensating transaction,Two-phase commit protocol,Serializability,Concurrency control,Computer science,Online transaction processing,Schedule (computer science),Computer network,Real-time computing,Database transaction,Distributed transaction,Distributed computing | Conference |
ISBN | Citations | PageRank |
978-1-4503-5028-0 | 0 | 0.34 |
References | Authors | |
23 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hua Fan | 1 | 14 | 6.04 |
Wojciech Golab | 2 | 210 | 17.22 |
Charles B. Morrey III | 3 | 0 | 1.35 |