Petroleum exploration seeks and ascertains oil and gas resources by using various exploration methods to understand subsurface geology. Traditional exploration follows straight survey lines, and the resulting data reflect only information in a two-dimensional plane beneath the line. That 2D form makes it hard to determine the presence of side reflections and to judge fault strike.

To address this, Sinopec adopted a new survey-line method, laying out a number of lines over a given surface area to obtain a 3D data volume of sufficient density. However, 3D raw data are large, diverse in form and mixed in type, posing a great challenge to back-end data storage. 

Sugon recently won the parallel-storage project of Sinopec Geophysical Research Institute. Following the customer’s exploration plan, the project carries out merged processing of four 3D survey areas, using a distributed storage system that is easy to expand, compatible and highly concurrent to handle the huge data those areas produce.

Easy expansion to meet PB-scale data challenges

In this project Sugon had to face a PB-scale data challenge from a 1,055-square-kilometer mining-rights block, helping the customer acquire seismic data in the block and thereby providing an important data basis for locating oil or natural-gas well sites.

To meet the project’s requirements for storage capacity, performance and security, Sugon used the ParaStor distributed storage system. ParaStor adopts a parallel architecture combining distributed technology, network communications and data management. It is a high-end storage system with proprietary intellectual property, designed for massive unstructured data. Sugon provided no less than 7,200 TB of raw storage capacity, able to form a unified resource pool with existing storage, a global single file system and a unified namespace, ensuring compatibility of storage expansion.

3D seismic data are hard-won. Sugon added 20 storage nodes directly in the existing management interface, keeping original index nodes unchanged, so that capacity of the existing storage pool could increase without stopping business or losing data. After nodes are added, directory capacity of existing storage increases automatically, and expansion is transparent to front-end applications.

Independent R&D and deep industry work: Sugon storage that understands geophysical exploration

With more than a decade of independent R&D, ultra-large-scale deployment experience and on-demand custom development, the Sugon ParaStor distributed storage system has won customer recognition. Facing the massive-data challenge of petroleum exploration, ParaStor has helped customers expand easily. Facing seismic data-processing software with as many as a thousand business modules, ParaStor analyzed and optimized I/O models for large amounts of random I/O, time to open ultra-large seismic data, working-directory listing and other layers, achieving optimal storage-layer performance for common modules of seismic processing software.

Sugon has for many consecutive years obtained directed-procurement qualifications from multiple geophysical-exploration units of PetroChina, Sinopec and CNOOC. With 20 years of deep understanding of industry application characteristics, Sugon will continue to optimize products and tune performance, address explosive growth in geophysical data storage, and provide more excellent practical support for the industry’s intelligent transformation.