智慧物流生态云平台计算能力测试数据集

该数据集用来测试平台主计算150M/s，全流程计算负载500万单情况下<5min；数据集通过智慧物流生态云平台获取，使用工具 Spark v6.3.2监测计算引擎（依托Impala）从磁盘获取数据（依托 Kudu）的平均速率265MB/s，大于150MB/s，使用Spark v6.3.2监测订单数据进入大数据，大数据抓取 12:15:54 到12:21:06之间的订单数据进行计算并分区域存储，四个分区在5分11秒内负载订单数据总和为8480000（records），按秒比例计算：8480000÷（5x 60+11）x（5x60） =8180064，5分钟内负载数据8180064（records）大于500万 通过人工审核处理数据。This dataset is designed to test that the platform's main computing throughput reaches 150 MB/s, and the full-process computing task can be completed within 5 minutes when handling 5 million orders. The dataset is collected via the Smart Logistics Ecosystem Cloud Platform. Using Spark v6.3.2 to monitor the computing engine (built on Impala), the average data retrieval rate from disks (backed by Kudu) is measured at 265 MB/s, which exceeds the 150 MB/s threshold. Additionally, Spark v6.3.2 is used to monitor the order data flowing into the big data platform, which grabs the order data between 12:15:54 and 12:21:06 for computation and regionalized storage. The total order data load across four partitions within 5 minutes and 11 seconds amounts to 8,480,000 records. Calculated via per-second proportional scaling: 8,480,000 ÷ (5×60 + 11) × (5×60) = 8,180,064. The order data load within 5 minutes is 8,180,064 records, which exceeds 5 million. The data is processed via manual review.

搜集汇总

背景与挑战

背景概述

该数据集用于测试智慧物流生态云平台的计算能力，验证平台在主计算速率150M/s和全流程负载500万单情况下能在5分钟内完成处理。数据集通过实际监测显示计算引擎平均速率达265MB/s，5分钟内负载订单数据超过500万条，数据量较小为10.19MB，来源于国家重点研发计划项目，聚焦于物流系统管理的信息处理。