Hello all :
Last week we try to increase the cache number for evaluating the stability of IoTDB. Now I will report the result to you: Start Time: 2019-06-05 08:26:12 End Time: 2019-06-06 13:50:26 Total elapse time: 105846 second Timeseries Num : 100,000 [1000 Devices * 100 Sersors] Loaded 20000000000points in 47857.383s with 4workers (mean rate 417908.34 points/s) Total Operations 200000; Latency(ms): Avg 950.4918, MiddleAvg 923.244, Min 785, Max 52390 Total error num is 0. /////////////////////////////////////////////////////////////////////////// ### client mode ### CLIENT_NUMBER=4 GROUP_NUMBER=10 DEVICE_NUMBER=1000 SENSOR_NUMBER=100 CACHE_NUM=1000 LOOP=200 [ limited by disk - 200G ] /////////////////////////////////////////////////////////////////////////// Next time we will try to increase the number of time series to 200,000[1000 Devices and 200 Sensors]. That’s all. Please feel free to contact me if you have any questions Thank you. Best Reagrds! FengQingxin 冯庆新 [[email protected]<http://hotmail.com/>] 大数据系统软件国家工程实验室 On 2019/05/19 12:35:23, Xiangdong Huang <[email protected]> wrote: > Hi all,> > > We have got a VM from ASF (iotdb-vm.apache.org), with 16GB memory, 4 cores> > and 200GB space.> > > Now Qingxin and I have started a long term test for evaluating the> > stability of IoTDB.> > The workload simulates 4 clients writing time series data as fast as> > possible. The data represents for 200 devices, while each device has 100> > sensors. The workload is generated by iotdb-benchmark[1].> > > > [1] https://github.com/thulab/iotdb-benchmark> > > Best,> > -----------------------------------> > Xiangdong Huang> > School of Software, Tsinghua University> > > 黄向东> > 清华大学 软件学院> >
