System Overview DrIM distributed tx with htm rdma a Target: OLTP workloads over large volume of data a Two independent components using htM&RDMA Transaction layer& memory store o low cost distributed TX Achieve over 5.52 million TXs/sec for TPC-C on 6 nodes Worker threads Transaction ayer key/value ops key/value ops Memory Store 品日口日
: Distributed TX with HTM & RDMA □ Target: OLTP workloads over large volume of data □ Two independent components using HTM&RDMA Transaction layer & memory store □ Low COST distributed TX − Achieve over 5.52 million TXs/sec for TPC-C on 6 nodes 11 System Overview key/value ops Transaction Layer Memory Store key/value ops Worker Threads DrTM
Agenda Transaction Layer Memory Storage Implementation Evaluation
Agenda Transaction Layer Memory Storage Implementation Evaluation
Challenge# 1: Restriction of HTM HTM is only a compelling hardware feature for single machine platform 口 Distributed tx cannot direct№ y benefit from it Some instructions system events(e.g network I/O) will unconditionally abort HTM transactions o Like any RDMA ops: READ/WRITE, CAS, SEND/RECV How to glue multiple HiM transactions together by rdma while preserving serializability 13
HTM is only a compelling hardware feature for single machine platform □ Distributed TX cannot directly benefit from it 13 Challenge#1: Restriction of HTM Some instructions & system events (e.g. network I/O) will unconditionally abort HTM transactions □ Like any RDMA ops: READ/WRITE, CAS, SEND/RECV How to glue multiple HTM transactions together by RDMA while preserving serializability?
Combining htm with 2Pl Using 2PL to accumulate all remote records prior to accesses in an htm transaction a transform a distributed tx to a local one a Limitation require advanced know ledge of read/write sets of transactions Worker Threads Transaction Layer key/value opsRDMA-key/value ops Memory Store I This is similar with prior work (e.g. Sinfonia Calv in) and the case for typical OLTP workloads 4
14 Combining HTM with 2PL Using 2PL to accumulate all remote records prior to accesses in an HTM transaction □ Transform a distributed TX to a local one □ Limitation: require advanced knowledge of read/write sets of transactions1 key/value ops key/value ops Transaction Layer Memory Store Worker Threads RDMA 2PL HTM 1 This is similar with prior work (e.g. Sinfonia & Calvin) and the case for typical OLTP workloads
DrTM's Concurrency Control Local tx vs local tx HTM Distributed tx vs distributed tx: 2PL Local x vs distributed tx: abort local Ix
15 DrTM’s Concurrency Control Local TX vs. Local TX: HTM Distributed TX vs. Distributed TX: 2PL Local TX vs. Distributed TX: abort local TX