vork has explored RDMA-based design of s et al.Low overh SIGMOD.2010. [29] asthe work 13o ng(ICPP), ly fo and is significantly more e efficient a high-pe ring 32 The tat ing the input tables l MDCC:200 7.CONCLUSION We argued that e CDE.2013 expe nts for OLTP h the net This is open 3可￥. rade-off to t lsTew8ogpicatioapmoidi simple abstractions to hide 8. REFERENCES IngniBand.In ICPP.2013 of Hadoop RPC with /doca/jesd-79-3d [43 J D19 6 olog 45A :/ A ification releas it2014 2013 0. C.Baik 4网 performance in 201 [so (14]C.Bin 网 Joins with Bloom nt locking fo or paralle high-spoce re for Compiling UDF-centri aidWtoa An ldea Who D roicct 1]A.Drago Memory.In NSD for determinism in databas 22A. 60q nline analytical processing on 61 utions in RAMCloud.In tions in multicore in-m MEe RoCE:An et-InfiniBan Story.HPC ning in paralleenabled transaction protocols, previous work has explored RDMA-aware join algorithms [26, 13]. Unlike our approach, the work in [26] still had the assumption that networks had limited bandwidth (only 1.25GB/s) and therefore streams one relation across all the nodes (similar to a block-nested loop join). In contrast, our RRJ join (as well as the work in [13]) is an extension of the state-of-the-art in-memory join algorithms for RDMA and is significantly more efficient than the RDMA join in [26] when comparing to the published numbers. Moreover, unlike from [13], our join supports effi- cient scale-out without re-partitioning the input tables. 7. CONCLUSION We argued that emerging fast network technologies ne￾cessitate a fundamental rethinking of the way we build dis￾tributed DBMSs. Our initial experiments for OLTP and OLAP workloads already indicated the potential of fully leveraging the network. This is a wide open research area with many interesting research challenges to be addressed, such as the trade-off between local vs. remote processing or creating simple abstractions to hide the complexity of RDMA verbs. 8. REFERENCES [1] www.jedec.org/standards-documents/docs/jesd-79-3d. [2] http://snowflake.net/product/architecture. [3] http://www.jcmit.com/memoryprice.htm. [4] Delivering Application Performance with Oracles InfiniBand Tech. http://www.oracle.com/technetwork/server-storage/ networking/documentation/o12-020-1653901.pdf, 2012. [5] Shared Memory Communications over RDMA. http://ibm.com/software/network/commserver/SMCR/, 2013. [6] Intel Data Direct I/O Technology. http://www.intel.com/ content/www/us/en/io/direct-data-i-o.html, 2014. [7] I. T. Association. InfiniBand Architecture Specification Release 1.2.1. http://www.infinibandta.org/, 2010. [8] I. T. Association. InfiniBand Roadmap. http://www.infinibandta.org/, 2013. [9] S. Babu et al. Massively parallel databases and mapreduce systems. Foundations and Trends in Databases, 2013. [10] P. Bailis et al. Eventual consistency today: limitations, extensions, and beyond. Comm. of ACM, 2013. [11] C. Balkesen et al. Multi-core, main-memory joins: Sort vs. hash revisited. In VLDB, 2013. [12] V. Barshai et al. Delivering Continuity and Extreme Capacity with the IBM DB2 pureScale Feature. IBM Redbooks, 2012. [13] C. Barthels et al. Rack-scale in-memory join processing using RDMA. In SIGMOD, 2015. [14] C. Binnig et al. Distributed snapshot isolation: Global transactions pay globally, local transactions pay locally. VLDB Journal, 2014. [15] M. Brantner et al. Building a database on S3. SIGMOD, 2008. [16] D. G. Campbell et al. Extreme scale with full sql language support in microsoft sql azure. In SIGMOD, 2010. [17] J. C. Corbett et al. Spanner: Googles globally distributed database. ACM TOCS, 2013. [18] A. Crotty et al. An Architecture for Compiling UDF-centric Workflows. In VLDB, 2015. [19] C. Curino et al. Schism: a Workload-Driven Approach to Database Replication and Partitioning. In VLDB, 2010. [20] D. J. DeWitt et al. The Gamma Database Machine Project. IEEE Trans. Knowl. Data Eng., 1990. [21] A. Dragojevic et al. FaRM: Fast Remote Memory. In NSDI, 2014. [22] A. J. Elmore et al. Squall: Fine-Grained Live Reconfiguration for Partitioned Main Memory Databases. In SIGMOD, 2015. [23] S. Elnikety et al. Database replication using generalized snapshot isolation. In SRDS, 2005. [24] F. F¨arber et al. The SAP HANA Database – An Architecture Overview. IEEE Data Engineering Bulletin, 2012. [25] M. Feldman. RoCE: An Ethernet-InfiniBand Love Story. HPC wire, 2010. [26] P. Frey et al. A spinning join that does not get dizzy. In ICDCS, 2010. [27] N. S. Islam et al. High performance RDMA-based design of HDFS over InfiniBand. In SC, 2012. [28] E. P. C. Jones et al. Low overhead concurrency control for partitioned main memory databases. In SIGMOD, 2010. [29] J. Jose et al. Memcached design on high performance RDMA capable interconnects. In Parallel Processing (ICPP), 2011. [30] A. Kalia et al. Using RDMA efficiently for key-value services. In SIGCOMM, 2014. [31] R. Kallman et al. H-store: a high-performance, distributed main memory transaction processing system. In VLDB, 2008. [32] D. Kossmann. The state of the art in distributed query processing. ACM Comput. Surv., 2000. [33] T. Kraska et al. MDCC: multi-data center consistency. In EuroSys, 2013. [34] J. Lee et al. SAP HANA distributed in-memory database system: Transaction, session and metadata management. In ICDE, 2013. [35] V. Leis et al. Morsel-driven parallelism: a NUMA-aware query evaluation framework for the many-core age. In SIGMOD. [36] J. J. Levandoski et al. High Performance Transactions in Deuteronomy. In CIDR, 2015. [37] Y. Lin et al. Middleware based data replication providing snapshot isolation. In SIGMOD, 2005. [38] Y. Lin et al. Snapshot isolation and integrity constraints in replicated databases. ACM Trans. Database Syst., 2009. [39] S. Loesing et al. On the Design and Scalability of Distributed Shared-Data Databases. In SIGMOD, 2015. [40] X. Lu et al. High-performance design of Hadoop RPC with RDMA over InfiniBand. In ICPP, 2013. [41] P. MacArthur et al. A performance study to guide RDMA programming decisions. HPCC-ICESS, 2012. [42] C. Mohan et al. Transaction Management in the R* Distributed Database Management System. In TODS, 1986. [43] J. K. Ousterhout et al. The case for ramcloud. Commun. ACM, 2011. [44] M. T. Ozsu. Principles of Distributed Database Systems. Prentice Hall Press, 3rd edition, 2007. [45] A. Pavlo. On Scalable Transaction Execution in Partitioned Main Memory Database Management Systems. PhD thesis, Brown University, 2014. [46] A. Pavlo et al. Skew-aware automatic database partitioning in shared-nothing, parallel OLTP systems. In SIGMOD, 2012. [47] O. Polychroniou et al. A comprehensive study of main-memory partitioning and its application to large-scale comparison- and radix-sort. In SIGMOD, 2014. [48] O. Polychroniou et al. Track join: distributed joins with minimal network traffic. In SIGMOD, 2014. [49] A. Pruscino. Oracle RAC: Architecture and performance. In SIGMOD, 2003. [50] A. Quamar et al. SWORD: scalable workload-aware data placement for transactional workloads. In EDBT, 2013. [51] V. Raman et al. DB2 with BLU acceleration: So much more than just a column store. In VLDB, 2013. [52] S. Ramesh et al. Optimizing Distributed Joins with Bloom Filters. In ICDCIT, 2008. [53] K. Ren, A. Thomson, and D. J. Abadi. Lightweight locking for main memory database systems. In VLDB, 2012. [54] W. R¨odiger et al. Locality-sensitive operators for parallel main-memory database clusters. In ICDE. [55] W. R¨odiger et al. High-speed query processing over high-speed networks. CoRR, abs/1502.07169, 2015. [56] R. Schenkel et al. Federated transaction management with snapshot isolation. In FMLDO. 1999. [57] M. Stonebraker et al. “One Size Fits All”: An Idea Whose Time Has Come and Gone. In ICDE, 2005. [58] H. Subramoni et al. RDMA over Ethernet - A preliminary study. In CLUSTER, 2009. [59] A. Thomson et al. The case for determinism in database systems. In VLDB, 2010. [60] C. Tinnefeld et al. Elastic online analytical processing on RAMCloud. In EDBT, 2013. [61] C. Tinnefeld et al. Parallel join executions in RAMCloud. In Workshops ICDE, 2014. [62] S. Tu et al. Speedy transactions in multicore in-memory databases. In SOSP, 2013. [63] E. Zamanian et al. Locality-aware partitioning in parallel database systems. In SIGMOD, 2015. 12