Search results for: Lanting%20Xia

Authors: Haitao Yang, Zhenjiang Ruan, Fei Xu, Lanting Xia

Abstract:

Industry data centers often need to sync data changes reliably and instantly from a large-scale of heterogeneous autonomous relational databases accessed via the not-so-reliable Internet, for which a practical generic sync middleware of low maintenance and operation costs is most wanted. To this demand, this paper presented a generic sync middleware system (GSMS), which has been developed, applied and optimized since 2006, holding the principles or advantages that it must be SyncML-compliant and transparent to data application layer logic without referring to implementation details of databases synced, does not rely on host computer operating systems deployed, and its construction is light weighted and hence of low cost. Regarding these hard commitments of developing GSMS, in this paper we stressed the significant optimization breakthrough of GSMS sync delay being well below a fraction of millisecond per record sync. A series of ultimate tests with GSMS sync performance were conducted for a persuasive example, in which the source relational database underwent a broad range of write loads (from one thousand to one million intensive writes within a few minutes). All these tests showed that the performance of GSMS is competent and smooth even under ultimate write loads.

Keywords: Heterogeneous massive data, instantly sync intensive writes, Internet generic middleware design, optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 382

Authors: Haitao Yang, Jianming Lv, Fei Xu, Xintong Wang, Yilin Huang, Lanting Xia, Xuewu Zhu

Abstract:

Given a fixed fund, purchasing fewer hosts of higher capability or inversely more of lower capability is a must-be-made trade-off in practices for building a Hadoop big data platform. An exploratory study is presented for a Housing Big Data Platform project (HBDP), where typical big data computing is with SQL queries of aggregate, join, and space-time condition selections executed upon massive data from more than 10 million housing units. In HBDP, an empirical formula was introduced to predict the performance of host clusters potential for the intended typical big data computing, and it was shaped via a regression approach. With this empirical formula, it is easy to suggest an optimal cluster configuration. The investigation was based on a typical Hadoop computing ecosystem HDFS+Hive+Spark. A proper metric was raised to measure the performance of Hadoop clusters in HBDP, which was tested and compared with its predicted counterpart, on executing three kinds of typical SQL query tasks. Tests were conducted with respect to factors of CPU benchmark, memory size, virtual host division, and the number of element physical host in cluster. The research has been applied to practical cluster procurement for housing big data computing.

Keywords: Hadoop platform planning, optimal cluster scheme at fixed-fund, performance empirical formula, typical SQL query tasks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 781