Therefore, this paper briefly introduces the development of data processing technology, data warehouse and data lake technology from the historical perspective of data technology, and leads to the combination of big data and blockchain technology, CATA.network. And then it introduces the basic concept of CATA.network and use cases.

The history of data processing technology

The first software system that supports data storage and computation was database that originated in the 1960s. Relational databases that emerged in the 1970s became the most popular data management systems that are still in use today. In the late 1980s, the data warehouse theory, which is specifically oriented to data analysis and decision making, was proposed and became the main tool to uncover the value of data for a long time to come.

In 1970, the research prototypes of relational databases, System R and INGRES, were introduced, both of which were designed for on-line transaction processing (OLTP) applications. The real available products for relational databases did not appear until 1980, DB2 and INGRES respectively.

By 1990, a new trend began to emerge: companies needed to collect data from multiple operational databases into a single data warehouse for business intelligence purposes. Despite the huge investment and limited functionality, companies that invested in data warehouses got a good ROI.

Since then, data warehouses have been supporting the business decision making process of major enterprises. The key technologies of data warehousing include data modeling, ETL technology, OLAP technology and reporting technology. At present, the main data warehouse product suppliers include Oracle, IBM, Microsoft, SAS, Teradata, Sybase, etc.

Around 2000, with the rapid development of the Internet, the volume and the complexity of data increased dramatically, and the demand for data processing speed increased. So the era of big data came. As a result, the NoSOL database for unstructured data emerged, and the distributed storage computing architecture that broke the bottleneck of single-computer storage and computing capacity became mainstream, Apache Hadoop became synonymous with big data technology, and MPP (Massively Parallel Processing) architecture also became popular at this time.

Around 2010, the arrival of the mobile Internet era further promoted the development of big data; and the demand for real-time interactivity led to the emergence of stream processing frameworks represented by Storm and Flink; and the demand for unified storage of different types of data gave rise to the concept of data lake. At the same time, with the in-depth application of cloud computing technology, the concept of cloud-native, which brings the advantages of resource intensification and application flexibility, has emerged, and big data technology has completed the transformation from private deployment to cloud deployment and then to cloud-native.

Starting from 2020, with the advancement of digital transformation in various industries and the frequent occurrence of data security incidents, the focus of the development of big data technology has evolved from a single focus on efficiency improvement to “efficiency improvement, business empowerment, security enhancement and circulation promotion”.

For more information, you can refer to:

Official Website: https://www.cata.network

Github: https://github.com/cata-network

Document: https://cata-1.gitbook.io/cata-docs/

Bitcointalk: https://bitcointalk.org/index.php?topic=5353832

Twitter: https://twitter.com/cata_blockchain

NFTs, as the hottest topic in the blockchain field at present, have strong demands on the functions of storage and query, which also motivates and guides the developments and designs of CATA project. CATA has designed a set of interaction protocols for Flow’s Cadence contract layer, enabling integration and full traceability of token transactions and data storage, providing a one-stop solution for NFT publishing, storage, management and distribution. It is fully compatible with Cadence’s resource-oriented capabilities, allowing Cadence developers to easily connect NFT contracts to storage facilities.

The current NFT scenario relies on blockchain networks (e.g., Flow or ETH) to build a marketplace, and the storage of metadata often relies on centralized storage solutions (e.g., AWS S3), or the content is stored on a decentralized storage protocol (such as IPFS), and then is associated with the on-chain marketplace through additional operations, which brings a bad impact on the user experience of NFT. The CATA storage protocol spans from the circulation layer on blockchain, the metadata layer, to the underlying file system layer, and covers the business logic of NFT data storage, the associations between the on-chain and off-chain world, and the entire process of token circulation. It provides a one-stop NFT experience. The Figure 2 illustrates the architecture of the CATA’s solution for NFT storage and the Figure 3 shows a simple example of the data structure.

With the collaboration of CATA, users can obtain persistent and secure storage services through the underlying CATA file system to preserve the raw data of NFT. In addition, the intermediate CATA-database layer can be used for flexible storage and retrieval of structured data, such as NFT metadata. With the metadata layer, NFT project parties can flexibly use and verify the data integrity on the chain, or build some tools outside the chain such as NFT search engine, thus greatly improving the user experience（e.g., All of the movie NFTs by Robert Downey Jr can be easily searched and reached through the CATA facilities.).

Parts of this article are quoted from the CATA Technical White Paper v0.2. For more information, you can refer to:

Official Website: https://www.cata.network

Github: https://github.com/cata-network

Document: https://cata-1.gitbook.io/cata-docs/

Bitcointalk: https://bitcointalk.org/index.php?topic=5353832

Twitter: https://twitter.com/cata_blockchain

The need for storage in a decentralized world is growing every day, and current decentralized storage solutions are still a long way from being truly productized.

Let’s turn our attention back to the storage facilities of Web 2.0. In the era of Web 2.0, people pay more attention to the delivery and sharing of information. Applications and various Web services have become more flexible and diverse. Storage facilities are not only referring to CD-ROMs, disks and other storage devices, but also towards the direction of application storage. Simply put, application storage should be a service that integrates storage devices and upper-level software functions. Due to the flexibility and scalability of the software system itself, it can hide the complexity of hardware cluster management and reduce the difficulty of the development of upper-layer applications. And software and hardware collaboration can eliminate bottlenecks to a certain extent and improve overall operational efficiency.

In addition, the product form of storage services includes not only disk writes and reads, but also a data management system built on top of the underlying storage, following various storage layer protocols (file system protocols, file transfer protocols, database standard protocols, etc.) to support upper layer applications. A simple centralized service architecture is shown in Figure 1.a that the underlying file system supports the upper Big Data System(BDS), which in turn provides flexible and scalable storage capacity directly to the upper computation systems and applications. Unfortunately, while mature file systems are emerging in the decentralized world, there does not exist a mature and practical Big Data system, which is exactly the role that CATA is expected to play, as shown in Figure 1.b.

The goal of CATA is to build a complete decentralized data storage system. At the same time, CATA will build a continuously growing open-source technical community for decentralized storage, inspiring more developers to create new storage tools and standard protocols to meet the diverse needs of dApps for data management.

In the first stage, CATA plans to build a decentralized query engine that can support structured data management to improve the flexibility of the existing decentralized file system. Meanwhile, as metadata management is one of the most urgent problems for today’s NFT applications, we decide to use NFT as the first scenario.

Inspired by the distributed serverless query engine, CATA adopts a design scheme that decouples storage and retrieval functionality. The bottom layer is a persistent decentralized file storage network built through the IPFS protocol. The file system will completely get rid of centralized services (e.g. centralized metadata store module), and instead completely rely on P2P communication. It is able to split files at a specific granularity and use content addressing to locate data, ensuring data persistence and availability. At the same time, we creatively docked Drill, a distributed query engine, on top of the decentralized file system and built CATA-database Layer. In this way, the whole decentralized storage system will provide scalable storage functionalities for the upper Blockchain Systems and dApps. For the moment, CATA has already supported common data formats such as json, basic SQL query capabilities and data export capabilities, as well as eliminated the centralized metadata store module.

Parts of this article are quoted from the CATA Technical White Paper v0.2. For more information, you can refer to:

Official Website: https://www.cata.network

Github: https://github.com/cata-network

Document: https://cata-1.gitbook.io/cata-docs/

Bitcointalk: https://bitcointalk.org/index.php?topic=5353832

Twitter: https://twitter.com/cata_blockchain

Development of blockchain

Satoshi Nakamoto’s revolutionary creation of the consensus protocol opened a new era of decentralized computing, called blockchain. People can freely contribute resources and receive benefits in a fair and open network, and the decentralized ledger will honestly record all decisions made in the network without privileged parties. The evolution of decentralized protocols is similar to early Internet protocols, but in essence they have achieved true decentralized autonomy, and a brand-new network for providing upper-level values has also been born.

Excitingly, since its birth, blockchain has gradually spawned various real-world application scenarios. From the first decentralized ledger, to decentralized finance, to NFT, to decentralized games, etc., the decentralized network has proven its practical value and immeasurable potential.

Introduction of CATA

CATA Network’s mission is to provide a general-purpose cloud infrastructure for data storage and management. Unlike the existing mainstream decentralized storage solutions (e.g. Filecoin, Arweave, etc.), CATA is not a specific storage module (e.g. file system or database). Instead, the plan is to build a more complete cloud storage system, or rather, a decentralized Big Data system, to provide more flexible, scalable, and practical data management services for the upper layer of decentralized applications.

The architecture of CATA

CATA built a decentralized file system network based on the IPFS protocol, implementing raw data sharding, distribution, storage and backup in this layer to meet the persistence and security requirements of data storage. However, CATA’s goals for the underlying storage layer are not limited to this; performance and cost metrics are also considered. In the future blockchain, there will be more decentralized applications, leading to larger and more complex data management requirements. Ensuring the scalability of the system and establishing a competitiveness comparable to centralized cloud storage is the core goal of the future CATA file system layer.

From the perspective of traditional storage system architecture, the file system typically only focuses on the organization and management of the data itself on the storage device, only providing the management of file operations to the upper layer. The application itself is often more complex and variable for data management needs, such as:

• Data Type (e.g., structured data, unstructured data);
• Data Lifecycle (e.g., long term persistent data, short term cache);
• Usage (e.g., OLAP, OLTP);
• Workload type (e.g., query-intensive, update-intensive);
• Other requirements (e.g, privacy protection, multiple backups).

Therefore, on top of the underlying file system layer, CATA will carry richer data management tools to build a complete storage infrastructure. In this way, decentralized applications can flexibly select and combine various storage services to meet their data management needs.

For now, CATA has implemented a decentralized metadata query engine to support Sql-based queries and custom retrieval of structured data for Dapp systems.

Future plan

CATA will complete the optimization of market incentive rules in the next phase to fully stimulate the value of personal storage resources (including storage, compute, bandwidth and other resources), so as to promote the development of a decentralized cloud storage market. In future releases, we will continue to integrate new decentralized storage tools into CATA. This is a long-term plan and will be the inevitable trend of decentralized storage development.

Parts of this article are quoted from the CATA Technical White Paper v0.1. For more information, you can refer to:

Official Website: https://www.cata.network

Github: https://github.com/cata-network

Document: https://cata-1.gitbook.io/cata-docs/

Bitcointalk: https://bitcointalk.org/index.php?topic=5353832

Twitter: https://twitter.com/cata_blockchain