Block Management¶
This article describes implementation details of management of blocks in the Aergo server. It is aimed at blockchain server developers.
Block generation¶
The block factory is responsible for generating blocks and part of the consensus engine.
The consensus engine manages when blocks are created on which nodes. AERGO uses DPoS by default. A new block producer is selected every second.
The process for creating blocks in the block factory is as follows.
- Select transactions from from mempool
- Perform a transaction one by one. If the execution is successful, the transaction is included in the block.
- If the block size and time limit are exceeded, the transaction is no longer included.
- Create blocks that contain executed transactions and add signatures to blocks.
The block factory sends the generated blocks to the chain service for propagation and storage.
Notify block¶
Blocks passed to the chain service are propagated to the connected peer nodes and stored in the database.
The chain service of a node that received the block verifies the block, executes the transactions, and adds the block to its own database.
Save block¶
The chain service stores the generated blocks in the DB for permanent storage. It also stores additional meta information to store connection information about the chain.
AERGO uses the Badger database as its permanent storage. Badger stores data in key/value form and supports the concept of transactions similar to a typical database.
Information stored in the DB is as follows.
The DB is managed in two separate categories, ChainDB and StateDB. This article describes the ChainDB used to store chain information.
Block info¶
This information is used to search for blocks in a chain using a hash. Typically, when a block is requested as a hash from another node, the Hash is searched for and responded to the block with a key.
| Key | Value |
|---|---|
| Hash of Block | serialized data of Block |
Chain info¶
Chain info is the information that indicates when a block is connected to the main branch.
If the main branch is not connected, that is, only block info is saved and the chain info is not stored.
You can search the blocks of every height of the main branch using the chain info. You can know the height of the top block of the chain.
Chain info is also used to determine whether the block belongs to the main branch or the sub branch.
| Key | Value |
|---|---|
| Number of Block | block hash |
| Latest block Key | latest block number of main branch |
Transaction Info¶
tx info is meta information of the transaction that was performed in the main branch. If you search the DB with transaction hash, you can see which block the transaction is stored in and where the transaction is stored in the block.
The contents of the transaction are not saved separately. Therefore, in order to know the contents of the actual transaction, you must search Block info using Block hash and index found in tx info.
| Key | Value |
|---|---|
| Transaction hash | (Block Hash, index of transaction in block) |