Indy Plenum - 术语表

• Node: A physical machine that is running one replica per protocol instance. The nodes are known to each other.
  节点：运行着针对于每个协议实例的一个副本的物理机器。节点是互相知道的。
• Instance: A logical construct that spans across all nodes, consisting of a replica from every node. Only one instance is the Master at a time. The rest are all called Backup
  实例：跨所有节点的一个逻辑结构，跟来自于每个节点的副本相同。每次只能有一个实例是 Master。剩余的节点被称为 Backup。
• Replica: A single node’s representative to an instance. A replica lives at the intersection of a node and a protocol instance. It lives on one node, and is dedicated to one and only one instance.
  副本：一个单一节点对于一个实例的表现。一个副本处在一个节点和一个协议实例的交叉点。它存在于一个节点之上，并且专门代表一个且仅一个实例。
• Pool: The collection of nodes running the consensus protocol
  Pool：运行着共识协议的节点的集合
• View: A configuration of the pool with designated leaders for each instance. Each view is known by a unique integer.
  View：包含了针对每个实例特定的 leader 的 Pool 的配置。每个 view 都含有一个唯一的数字值。
• View change: Process of choosing a new primary replica for each protocol instance, when the current master’s throughput or other performance metrics are lower by certain thresholds than that of other protocol instances
  View change：在当前的 master 的吞吐量或者其他的性能指标相比较于其他的协议实例降低到某个临界值的时候，为每个协议实例选择一个新的 primary 副本的流程
• Election: Process to decide which replica in a protocol instance would be the primary
  选举：选择在一个协议实例中哪个副本会作为 primary 的流程
• Request: A message sent by the client to write to or query from the ledger. Each request has a field called operation which is the payload of the request. The operation contains a field type which indicates the intent of the request.
  请求：由客户端发出的从账本上读取或者写入的一个消息。每个请求都含有一个名为 operation 的字段，它是请求的 payload。这个 operation 包含了一个说明该请求意图的 type 字段。
• Request Digest: A sha256 hash of the serialised request data.
  请求摘要：序列化后的请求数据的一个 sha256 哈希值
• Request ordering: Successful completion of consensus process over a request.
  请求排序：对于一个请求的共识流程的成功完成
• Transaction: A write request successfully processed (ordered) by the pool. Each transaction has a field called type which indicates the intent of the transaction.
  交易：由 pool 成功处理 (排序) 的一个写入请求。每个交易含有一个称为 type 的字段，它表示了交易的意图。
• Ledger: An ordered log of transactions. Each node hosts several ledgers which serve different purposes. Each ledger is uniquely identified by a ledger id. Each correct node hosts the same ledger as any other correct node. The ledger assigns each transaction a unique monotonically increasing positive integer called sequence number.
  账本：一个交易的有序 log。每个节点存储着具有不同目的的不同账本。每个账本由 ledger id 唯一标识。每个正确的节点存有跟其他任何节点都一样的账本。账本为每个交易分配了一个唯一的单项递增的数字值，称为顺序数值 sequence number。
• State: A projection of the ledger. Exposes a key-value store like API, can provide merkle proof of presence of keys with values. Currently uses a Merkle Patricia Trie under the hood.
  State: 账本的一个投影。通过 API 暴漏了一个 key-value 存储，能够提供表示 keys 和 values 的 merkle 证明。当前使用的是 Merkle Patricia Trie
• Catchup: The process of a node syncing a ledger with other nodes. Used when a node starts or during a view change or it finds itself lagging behind others.
  Catchup：一个节点同其他节点同步账本的过程。当节点启动或者在一个 view change 的过程中，或者当节点发现它落后于其他节点的时候会使用。
• Ledger Manager: Contains logic for catchup. Allows registering ledgers to be synced and callbacks to be called at different events.
  账本管理者：包含了 catchup 的逻辑。允许在不同的事件中注册的账本能够被同步以及回调调用。
• Request Handler: A class containing the processing logic of requests of certain type. For write requests, updates the ledger and might update state too.
  请求 Handler：一个包含针对特定 type 请求的处理逻辑的类。对于写请求，会更新账本并可能也会更新 state。
• Client: Generic logic for sending requests to node(s), handling retries, acknowledgments, replies, etc. Allows registering callbacks that are called once a request has been successfully processed by the pool.
  客户端：通用的逻辑，来向节点发送请求，处理重试，接受，回复等。允许注册回调函数以便当一个请求由 pool 成功处理之后来调用。
• Wallet: Stores private keys and other secrets. Also contains code to act on successfully processed requests.
  钱包：存储私钥以及其他的隐私信息。也包含了当成功处理请求之后的要做的动作的代码。