PuppetDB是一个傀儡的数据仓库，它管理所有平台生成的数据存储和检索。Currently, it stores catalogs and facts; in future releases, it will expand to include more data, like reports.目前，它存储目录和事实，在将来的版本中，将扩大到包括更多的数据，如报告，。

So far, we've implemented the following features:到目前为止，我们已经实现了以下功能：

Fact storage其实存储

Full catalog storage全目录存储

Containment edges遏制边缘

Dependency edges依赖边缘

Catalog metadata目录元数据

Full resource list with all parameters所有参数全部资源列表

Node-level tags节点级标签

Node-level classes节点级班

REST Fact retrieval REST事实检索

all facts for a given node对于一个给定节点的所有事实

REST Resource querying REST资源查询

super-set of storeconfigs query API超集查询API storeconfigs

boolean operators布尔运算符

can query for resources spanning multiple nodes and types可以查询跨越多个节点和类型的资源

Storeconfigs terminus storeconfigs总站

drop-in replacement of stock storeconfigs code下拉更换股票storeconfigs代码

export resources出口资源

collect resources收集资源

fully asynchronous operation (compiles aren't slowed down)完全异步操作（编译没有放慢）

much faster storage, in much less space更快的存储，在少得多的空间

Inventory service API compatibility库存服务API兼容性

query nodes by their fact values他们其实值的查询节点

drop-in replacement for Puppet Dashboard and PE console inventory service下拉更换木偶仪表板和PE控制台库存服务

Components组件

PuppetDB consists of several, cooperating components: PuppetDB几个，合作的组成部分包括：REST-based command processor基于REST的命令处理器　PuppetDB uses a CQRS pattern for making changes to its domain objects (facts, catalogs, etc). PuppetDB使用域对象（事实，目录等）的变化1 CQRS格局。Insteadof simply submitting data to PuppetDB and having it figure out theintent, the intent needs to be explicitly codified as part of theoperation.只需提交数据PuppetDB和它搞清楚的意图，而不是意图必须明确编纂作为行动的一部分。This is known as a “command” (eg “replace the current facts for node X”).这就是所谓的“命令”（如“节点X取代目前的事实”）。

Commandsare processed asynchronously; however, do our best to ensure that oncea command has been accepted, it will eventually be executed.命令是异步处理，然而，竭尽所能，以确保一次命令已被接受，它最终将被执行。Ordering is also preserved.还保存着排列。To do this, all incoming commands are placed in a message queue which the command processing subsystem reads from in FIFO order.要做到这一点，所有传入的命令被放置在消息队列中的命令处理子系统读取FIFO顺序。

Submission of commands is done via HTTP.通过HTTP提交的命令。See the API spec in the navigation sidebar for complete documentation.看到完整的文档，在导航栏的API规范。There is a specific required wire format for commands, and failure to conform to that format will result in an HTTP error.有一个特定的命令所需的电线格式，不符合该格式将导致HTTP错误。

Storage subsystem存储子系统

Currently, PuppetDB's data is stored in a relational database.目前，PuppetDB的数据存储在关系数据库中。There are two supported databases:支持的数据库有两个：

An embedded HSQLDB.一个嵌入的HSQLDB。This does not require a separate database service, and is thus trivial to setup.这并不需要一个单独的数据库服务，因而是微不足道的设置。This database is intended for proof-of-concept use; we do not recommend it for long-term production use.这个数据库被用于概念证明使用，我们不建议长期生产使用。

PostgreSQL PostgreSQL的

There is no MySQL support, as MySQL lacks support for recursive queries (critical for future graph traversal features).有没有MySQL的支持，因为MySQL缺乏支持递归查询（未来图的遍历功能的关键）。

REST-based retrieval基于REST的检索

Read-only requests (resource queries, fact queries, etc.) are done using PuppetDB's REST APIs.使用PuppetDB的REST API的只读请求（资源查询，其实查询等）完成。Each REST endpoint is documented in the API spec; see the navigation sidebar.每个REST端点被记录在API规范;看到导航栏。

Remote REPL远程的REPL

For debugging purposes, you can open up a remote clojure REPL and use it to modify the behavior of PuppetDB, live.出于调试目的，你可以打开一个远程的Clojure 的REPL，并用它来 修改的PuppetDB，现场的行为。

Vim support would be a welcome addition; please submit patches! Vim的支持将是一个值得欢迎的除了请提交补丁！