The future public chain must be multi chain ecology


The future public chain must be multi chain ecology
Whenever you see the blockchain project mentioned in the white paper, your first reaction is to express doubt. It’s not that it’s impossible at all, but that it’s necessary to take high TPS as a selling point, and to put it so high, it’s just that there is no it common sense. How bold do you think people are and how productive the land is. You can only bluff the layman. Million TPS, who can do it? Alibaba can do it! Let’s first take a look at how Alibaba handles the high concurrency transaction requests on the night of the 11th National Day every year.

In the figure below, from top to bottom, we can see that Alibaba allocates transaction requests at many levels

1、 DNS polling is used to realize the load balance of the machine room. Multiple IP addresses are configured in the DNS server, and each IP address corresponds to the virtual IP in different computer rooms.
2、 LVS and F5 are load balancing solutions that work in the fourth layer of the network. LVS is software that runs in the kernel state of the operating system and can forward TCP requests or higher level network protocols.
3、 The network servers are deployed on multiple servers, and the requests are distributed evenly to each server by using the reverse proxy software (nginx).
4、 According to the business sector to divide the application code, make different applications, such as Taobao, tmall and so on
5、 The enterprise service bus (ESB) is introduced to shield the access difference of service interface.
6、 The technology of containerization is introduced to realize running environment isolation and dynamic service management. So in the evening of double 11, it is very convenient to deploy more servers to support applications.
7、 The function of reuse is separated from microservice, so we can see that there are user management service, payment service, order service, etc.
8、 The technology of distributed file system, NoSQL database and search engine is introduced to reduce the pressure of database and improve the efficiency of data call.
9、 Inside the database, it also adopts the database read-write separation, dividing the database according to the business, and splitting the large table into small tables.

The original text can be read here:…

The future public chain must be multi chain ecology
The future public chain must be multi chain ecology
So, see, even Alibaba relies on a variety of multi-level, multi process, distributed storage + computing, and a large number of servers to achieve transactions of more than one million levels per second. How difficult it is to realize such a huge amount on a decentralized blockchain. A simple transaction and a multi-step process with complex business logic are more meaningless than TPS.

There is also a common problem in the project, which is to solve the well-known blockchain Impossible Triangle. What is the Impossible Triangle of blockchain? That is, the security, efficiency and decentralization of blockchain can not be achieved at the same time. What I want to say is that in a blockchain, it is impossible to say that it is impossible, and there is no need to change the concept to say that it is possible. If there is a possibility, it must be on the basis of ensuring security and decentralization, to achieve efficient transaction diversion. Therefore, the solution is nothing more than the following: fragmentation, layering, side chain and sub chain. Let’s call it multilayer structure.

Therefore, today we will study the realization or claim to realize the multi-layer architecture of the public chain.

It used to be a star project. Its main highlight is fragmentation. Let’s talk about the segmentation first. Slicing is actually a traditional database technology, which divides a large database into smaller, faster and easier to manage parts. Each part is called a shard, which supports parallel processing. So slicing can greatly improve the data processing ability. In the blockchain, fragmentation is to divide all transactions on the network into different groups, and each group is a slice. The slices composed of various nodes are processed at the same time, which disperses the work to each slice and speeds up the processing speed. It is divided into three types: state fragmentation, transaction fragmentation and network fragmentation. The so-called network fragmentation is to allocate the physical nodes in the network to different partitions, and build a separate network to support the upper transaction state; transaction fragmentation is to allocate transactions to different partitions for consensus, and the whole network can package and verify different transactions at the same time; status fragmentation is to store the complete account information in each partition, and each partition has its own status Self maintenance part of the account information, state fragmentation involves cross fragmentation data synchronization, cross slice transactions, in the three kinds of fragmentation is the most complex one.

Zilliqa adopts the pow consensus to realize transaction fragmentation rather than state fragmentation. A small number of nodes are replaced periodically in each partition, and pbft is used to achieve consensus in the partition.

Dr. Quan’s founding team, simple and clear way of improvement, pragmatic publicity, and indeed achieved nearly 1000 transactions per second in the test environment. So this project was very popular with investors. The first blog I posted on Jane’s book was to introduce it. It’s a pity that TPS hasn’t been more optimized since it went online, and its comparative advantage with EOS and other blockchains adopting dpos consensus is not obvious. And the founders left one after another. The outlook for zilliqa is not optimistic.

Ethereum 2.0
Ethereum was the first to propose the concept of fragmentation. As early as 2016, there have been proposals for fragmentation. However, Ethereum is too large, and many security factors need to be considered for upgrading. So the road to upgrading is long. It was not until Istanbul was upgraded in December this year that the road of 2.0 was opened. However, it is estimated that it will be several years before the real fragmentation can be realized. However, the fragmentation mode of Ethereum is worth exploring. It will be implemented in several steps.

The first step is to launch the beacon chain. This is the core of the eth2.0 ecosystem and the root of the security and validation of other segments. After the beacon chain is deployed, POW / POS hybrid mechanism will be used for equity certification.

The second step is to implement state slicing, and establish the communication mode between slicing and between slicing and beacon chain. The users are divided into different chips to realize state partition. Then the state is finally authenticated on the beacon chain.

The third step is to implement eth2.0 which is fully partitioned and integrated. Fragmentation will be upgraded from “basic data producer” to “full function chain”. And the introduction of a new Ethereum virtual machine ewasm.

In fact, there are still many uncertainties in Ethereum’s plan for fragmentation. For the time being, there are only some plans and proposals, such as:

In the management of node fragmentation, beacon chain needs to provide randomness and unpredictability by using random number protocol and verifiable delay function to prevent potential attackers from trying to predict which partition their nodes will be allocated to.

We need to design a special protocol to ensure efficiency and security. Fragmentation also needs to know when it is appropriate to communicate with each other and only when it is needed.

As the busiest blockchain, Ethereum moves the whole body at once. Every upgrade requires careful planning and the cooperation of all parties. Therefore, although I firmly believe that eth2.0 will be successfully implemented, and will help Ethereum achieve large-scale expansion and adoption, the market can not afford to wait. In fact, many new public chain projects have realized the fragmentation theory proposed by Ethereum.

Harmony realizes state fragmentation, that is to say, the ledger data is also processed by fragmentation. It uses the theory of eth 2.0 for reference. The fragmentation architecture includes one beacon chain and multiple fragmentation chains.

Harmony uses the optimized pbft in slicing. The optimization part is to add BLS signature, so that the message volume changes from O (n * n) to o (n).

Harmony also provides randomness for fragmentation by running distributed random number protocol and verifiable delay function; beacon chain adopts POS, accepts the mortgage token of node, and finally confirms the state of sub chain.

Top Network
Top network uses network fragmentation. It designs a three-layer network. The upper layer is a beacon network similar to the beacon chain of eth2.0, which manages the fragment; the fragment belongs to the consensus network, which processes the transaction; in addition, there is an audit network, which audits the transaction that has been verified by the fragment, so as to prevent the transaction within the fragment from being tampered. Of course, there are also some innovations, such as adopting the simplified DAG idea and making consensus at the transaction level rather than based on blocks. Therefore, the transaction confirmation is fast. However, Mingming admits that a single TPS is just like 500, and does not actually support any high concurrency applications. However, in all kinds of publicity, he says that the single chain processing capacity exceeds 300000 TPS, and the marketing is very unprofessional.

MOAC should be the first public chain to put forward the parent-child chain architecture, and the main network will be launched in 2018. Later, the architecture of MOAC was renamed foundation chain and application chain. The foundation chain provides the source of security and verification for the application chain. The application chain is generated from the basic chain and shares the account information with the basic chain and other application chains, but it is an independent blockchain, running independent transactions and retaining the account status. Then refresh the status to the base chain by refresh periodically. This implementation method was used for reference by many public chains later.

Another innovation of MOAC is that the application chain can support multiple consensus, and users can choose consensus mechanism according to their own business needs. This is a further upgrade than simple fragmentation, because this kind of support for multiple consensus can enable MOAC to support a variety of completely different applications, so MOAC is a real blockchain for commercial applications. The distributed storage project file storm is an application chain implemented on MOAC.

According to the development roadmap of MOAC, it will support multi tier architecture in the future. Any application chain can become the basic chain of a new application chain. Such an architecture can support a business empire like Alibaba that needs high throughput and high concurrency in the future.

Because of being selected as a blockchain partner by Huoyuan, and raising $25 million in compliance financing in the United States, nervos has become a popular fried chicken in the public chain. The architecture of layer 1 and layer 2 proposed by them is no different from the way of using beacon chain and fragmentation chain to expand capacity. But it’s obvious that the project nervos has incorporated a lot of engineering theory and open thinking. Layer 2 of nervos can be an independent blockchain, a fragment, a state channel, or even a DAPP. Layer2 develops freely. Only when you have to use security issues, you can call layer1. Go out of the simple idea of TPS expansion and increase the diversity of layers from the perspective of practicability. On the contrary, nervos will become more scalable. This is worth learning.

Cosmos and Polkadot
Finally, let’s talk about these two projects. They implement cross chain in different ways. In fact, it is also to connect different chains into an ecology. In the future, many public chains will support cosmos or Polkadot protocol to exchange data and value with other public chains.

From the above public chain analysis, it is not difficult to see. Hierarchical architecture, or multi chain ecology, is the general trend of public chain in the future. Moreover, in order to support complex business logic, different applications have different requirements for data storage and exchange, and different computing performance, the future public chain must support complex architecture, diverse consensus, and reasonable allocation of storage and computing resources, but through the standard API interface, it is more convenient for end users to use.

In addition, there is another part that is often ignored, which is the economic incentive of blockchain to nodes. Whether the economic model is reasonable or not is the key to the long-term operation of the public chain. The complex architecture will bring about the difference of node rewards, and if the difference is too large, it will affect the security of the blockchain. Most of the public chains have changed from POW to POS, but the consensus of POS has not experienced the verification of time like pow. How should the encrypted assets issued by users on the public chain interact with the original certificate of the public chain, and what kind of economy will be formed in the future. In addition, can the alliance chain without money coexist with the public chain with money. These are all new areas that need to be further developed.

A new attempt of public chain
Therefore, considering the advantages and disadvantages of all kinds of segmented public chains, the author puts forward some new ideas here

  1. The hybrid consensus mechanism of pow + POS is adopted on the beacon chain / base chain, which not only ensures the safety of the chain, but also prevents miners from obtaining all rewards through arms competition through stacking.
  2. The POS node can be a fragment chain / application chain node.
  3. DAPP can use base chain or independent application chain. If the transaction volume of an application increases sharply, multiple fragmentation chains can be used automatically.
  4. Each application can choose to generate the native token on the chain. The original token and the basic chain token can be used through the same API to support wallet and exchange.
  5. The code is completely open source, join Dao’s chain governance.

The core of blockchain is consensus. The first thing to reach a consensus is the technology developers. Welcome blockchain experts and programmers to discuss these new ideas with us. If we can reach a consensus and then develop together, the public chain will have a great market. If we can make a technological breakthrough together and achieve an efficient blockchain system like Alibaba, we will certainly be able to carry the heavy task of enabling the real economy.

Why is filestorm a public chain
When it comes to this, some people may ask. File storm is a distributed storage application based on blockchain. Just store it well. Why go back to the public chain? This is because storage alone cannot make full use of the value of data. It also needs reasonable data sharing and efficient calculation. The three elements of computing are storage, computing and network. If blockchain wants to become a real supercomputer, it must take these three into account. The file storm blockchain has realized the storage. In addition to the blockchain, the integration of computing in the future will enable the data to generate more value and empower the industry.

File storm is an application chain implemented on MOAC and an extension of MOAC. MOAC supports multi-layer chain architecture. File storm hopes to add its innovation to the multi tier architecture of MOAC application chain, and create a new distributed storage Multi Chain ecosystem. Provide better blockchain services and storage services for various decentralized applications.

The State encourages independent innovation of blockchain technology. It has accumulated a lot of experience in docking blockchain applications, which can be used to improve itself and strive for a breakthrough.

For example, dostorm’s dpos consensus can be easily transformed into an alliance chain, and distributed storage can be regarded as an important part of blockchain-as-a-service. So that filestorm can serve more government and enterprise units and small and medium-sized enterprises. With the public chain thinking to do alliance chain, can better promote the average distribution of social value, and fair competition in the industry.

Author: blockchain_ Brother Xian