Kovan Testnet

Kovan is a Proof of Authority (PoA) publicly accessible blockchain for Ethereum; created and maintained by a consortium of Ethereum developers, to aide the Ethereum developer community.

This new testnet using Parity (an Ethereum client developed by Parity Technologies, FKA Ethcore) to provide a stable, secure testnet environment for Ethereum developers, due to the instability of the existing Ropsten testnet.

On 24th Feb 2017, Ropsten was under a denial-of-service attack (“spam attack”). Average block propagation time has since slowed to a crawl as a large miner has decided to deploy several zero-value high-gas transactions to spam the test network continually. The attacker’s intentions are unknown, but the result is that Ethereum developers who rely on Ropsten no longer have a stable public testing environment to deploy and test their smart contract code prior to deploying to production on the mainnet chain.

The use of Proof-of-Work (PoW) for testnets presents a fundamental game theoretical problem: the only significant economic incentive to mine on testnets using dedicated GPU resources is to launch an attack and reduce stability and the viability of the testnet (and thus hamper development for the mainnet chain).

Parity supports a PoA consensus engine to be used with Ethereum Virtual Machine (EVM) based chains. PoA is a replacement for PoW, which can be used for both public and private chain setups. There is no mining involved to secure the network with PoA, and relies on trusted ‘Validators’ to ensure that valid transactions are added to blocks, processed and executed by the EVM faithfully.

Because mining does not occur on our proposed public test net, malicious actors are prevented from acquiring testnet Ether, solving the spam attack that Ropsten faced.

There is no difference in the way that contracts are executed compared to PoW chains, so developers can test their contracts and user interfaces before deploying to the mainnet in a more reliable and convenient environment.

More information about PoA can be found here.

Aside from solving the critical stability issue of the ongoing Ropsten “spam attack”, there are additional benefits for providing a public consortium PoA network.

• Shorter block times, allowing for more rapid deployment, testing and iteration
• Reduced overall maintenance costs (no cpu-intensive mining is required)

A select group of trusted parties (“Consortium”) will be responsible for maintaining a cluster of Ethereum nodes running PoA to verify blocks.

To ensure a sufficient degree of trust and redundancy, a minimum number of Consortium members should be recognised as being trusted members of the Ethereum/Blockchain community. Servers should not be controlled by a single entity, but run individually by the consortium’s companies, preferably with servers located in multiple regions.

Formal process to be confirmed. Currently using gitter for decision making. The Parity codebase (and thus Kovan) is ultimately controlled by Parity Technologies.

• 4 second block time
• Parity AuthorityRound PoA consensus mechanism
• Use --force-sealing
• Homestead rules with possible Metropolis & beyond additions;

A secure “Faucet” service will be provided to allow for verified (non-malicious) developers to acquire testnet Ether. It is important that the distribution of testnet Ether is available but is also rate-limited, so as to be not available in large amounts to non-trusted parties (to prevent spam attacks).

The faucet will have a web-based interface, and will require some level of verification (to be determined), but could include:

• Github User Verification (OAuth)
• SMS Verification (already developed by Parity)
• Manual KYC from consortium members

It is to be determined whether or not the faucet service will require manual approval from the consortium validators, with one or more validators approving requests for testnet Ether. This could be achieved using an admin backend for approving requests optionally, with an on-chain multi-sig wallet to prevent an individual validator from ‘going rogue’.

Another approach could leverage the economic properties of the Ethereum mainnet, by automatically granting testnet Ether to users who send mainnet Ether into a specific contract (which could in turn be used to fund maintenance of the Kovan testnet). This approach would grant privacy benefits to developers, whilst still creating an economic barrier to entry that prevents large amounts of testnet Ether from being accrued.

An important requirement for testnet developers is to be able to easily verify that transactions have been processed using a third party interface. The most popular blockchain explorer for Ethereum is Etherscan, which currently provides an explorer interface for both the Ethereum mainnet and Ropsten testnet. Etherscan will also provide this service and also a set of API endpoints for the Kovan testnet.

Any developer, including non-consortium-members, can interact with the blockchain by running a local Parity node using a specific configuration.

Additionally, a number of public JSON-RPC endpoints will be provided for deploying Smart Contracts and interacting with the network, where developers can use “zero-wallets” for signing transactions, providing further convenience for any given dev environment - this removes the requirement of running a local node, and is useful for mobile and web-based applications.

Ideally, a scalable load-balanced service (such as Infura) should be used to ensure uptime and reliability.

It is critical that validators have correctly-configured PoA validation nodes running in reliable and redundant way. Various deployment options will be provided to validators:

• Docker Images (TBC)
• Manual Configuration/Setup Documentation

Etherscan (Block Explorer Provider / PoA Validator)

Parity Technologies (PoA Validator / Parity Client Developer)

DigixGlobal (PoA Validator)

Attores (PoA Validator)

Maker / Dapphub (PoA Validator)

OneBit / TenX (PoA Validator)

Aurel Iancu, Prominent Miner / Contributor, Ethereum Advocate (PoA Validator)

Melonport (PoA Validator)

GridSingularity (PoA Validator)

Nivaura (PoA Validator)

Additional consortium members could be added to provide increased trust and provide development resources for the Faucet service.

The listed PoA Validator entities will be (at least) responsible for maintaining validation nodes, and ideally with redundancy on other services to prevent critical failure if is in some way compromised. There will also be to-be-determined development responsibilities from these parties with regards to providing JSON RPC nodes and the Faucet service.

Etherscan: 0x00D6Cc1BA9cf89BD2e58009741f4F7325BAdc0ED

Parity: 0x0010f94b296a852aaac52ea6c5ac72e03afd032d

DigixGlobal: 0x007733a1FE69CF3f2CF989F81C7b4cAc1693387A

Attores: 0x00427feae2419c15b89d1c21af10d1b6650a4d3d

Maker: 0x00E6d2b931F55a3f1701c7389d592a7778897879

TenX (OneBit): 0x4Ed9B08e6354C70fE6F8CB0411b0d3246b424d6c

Aurel: 0x00e4a10650e5a6D6001C38ff8E64F97016a1645c

Melonport: 0x0020ee4Be0e2027d76603cB751eE069519bA81A1

GridSingularity: 0x00a0a24b9f0e5ec7aa4c7389b8302fd0123194de

Nivaura: [will join in first HF]