NGAS-Based Multi-site Observation Data Storage and Synchronization Method

The Square Kilometre Array (SKA) will have ultra-high sensitivity, ultra-fast survey speed and wide field of view, resulting in super-massive raw observation data. Massive data synchronization/delivery between the SKA observatories and regional data centers in various countries is a difficult problem in the current SKA construction. The Next Generation Archive System (NGAS) used by SKA precursor is inefficient and has insufficient performance in the application measurement. In this paper, we proposed a ZeroMQ-based data storage and synchronization method. By adopting more efficient asynchronous messaging mechanism to realize synchronous data delivery, it can avoid the limitations of the HTTP protocol adopted by the NGAS. Experimental results show that the new method is nearly 40 times faster than the original method used by NGAS in terms of average data archival/storage efficiency, which can basically meet the requirements of the full-speed transmission of 10GB bandwidth and has achieved a better use effect.