Git Product home page Git Product logo

firefly-fabconnect's Introduction

firefly-fabconnect

A reliable REST and websocket API to interact with a Fabric network and stream events.

Architecture

High Level Components

high level architecture

Objects and Flows

objects and flows architecture kafkal handler architecture

The component provides 3 high level sets of API endpoints:

  • Client MSPs (aka the wallet): registering and enrolling identities to be used for signing transactions
  • Transactions: submit transactions and query for transaction result/receipts
  • Events: subscribe to events with regex based filter and stream to the client app via websocket

Getting Started

After checking out the repo, simply run make to build and test.

To launch, first prepare the 2 configurations files:

  • sample main config file:
{
  "maxInFlight": 10,
  "maxTXWaitTime": 60,
  "sendConcurrency": 25,
  "receipts": {
    "maxDocs": 1000,
    "queryLimit": 100,
    "retryInitialDelay": 5,
    "retryTimeout": 30,
    "leveldb": {
      "path": "/Users/me/Documents/ff-test/receipts"
    }
  },
  "events": {
    "webhooksAllowPrivateIPs": true,
    "leveldb": {
      "path": "/Users/me/Documents/ff-test/events"
    }
  },
  "http": {
    "port": 3000
  },
  "rpc": {
    "useGatewayClient": true,
    "configPath": "/Users/me/Documents/ff-test/ccp.yml"
  }
}
  • the standard Fabric common connection profile (CCP) file that describes the target Fabric network, at the location specified in the main config file above under rpc.configPath. For details on the CCP file, see Fabric SDK documentation. Note that the CCP file must contain the client section, which is required for the fabconnect to act as a client to Fabric networks.

Use the following command to launch the connector:

./fabconnect -f "/Users/me/Documents/ff-test/config.json"

API Specification

The API spec can be accessed at the endpoint /api, which is presented in the Swagger UI.

swagger ui

Hierarchical Configurations

Every configuration parameter can be specified in one of the following ways:

  • configuration file that is specified with the -f command line parameter. this is overriden by...
  • environment variables that follows the naming convention:
    • given a configuration property in the configuration JSON "prop1.prop2"
    • capitalized, exchanging . with _, then add the FC_ prefix
    • becoming: FC_PROP1_PROP2
    • this is overriden by...
  • command line parameter with a naming convention that follows the same dot-notaion of the property:
    • given "prop1.prop2"
    • the command line parameter should be --prop1-prop2 or a shorthand variation

Support for both Static and Dynamic Network Topology

There is support for using a full connection profile that describes the entire network, without relying on the peer's discovery service to discover the list of peers to send transaction proposals to. A sample connection profile can be seen in the folder test/fixture/ccp.yml. This mode will be running if both rpc.useGatewayClient and rpc.useGatewayServer are missing or set to false.

There is also support for using the dynamic gateway client by relying on the peer's discovery service with a minimal connection profile. A sample connection profile can be seen in the folder test/fixture/ccp-short.yml. This mode will be running if rpc.useGatewayClient is set to true.

Support for server-based gateway support, available in Fabric 2.4, is coming soon.

Structured Data Support for Transaction Input with Schema Validation

When calling the POST /transactions endpoint, input data can be provided in any of the following formats:

  • in the "traditional" array of strings corresponding to the target function's list of input parameters:
POST http://localhost:3000/transactions?fly-sync=true&fly-signer=user001&fly-channel=default-channel&fly-chaincode=asset_transfer
{
    "headers": {
        "type": "SendTransaction"
    },
    "func": "CreateAsset",
    "args": ["asset204", "red", "10", "Tom", "123000"]
}
  • provide a payloadSchema property in the input payload headers, using JSON Schema to define the list of parameters. The root type must be an array, with prefixItems to define the sequence of parameters:
POST http://localhost:3000/transactions?fly-sync=true&fly-signer=user001&fly-channel=default-channel&fly-chaincode=asset_transfer
{
    "headers": {
        "type": "SendTransaction",
        "payloadSchema": {
            "type": "array",
            "prefixItems": [{
                "name": "id", "type": "string"
            }, {
                "name": "color", "type": "string"
            }, {
                "name": "size", "type": "integer"
            }, {
                "name": "owner", "type": "string"
            }, {
                "name": "value", "type": "string"
            }]
        }
    },
    "func": "CreateAsset",
    "args": {
        "owner": "Tom",
        "value": "123000",
        "size": 10,
        "id": "asset203",
        "color": "red"
    }
}
  • when using payloadSchema, complex parameter structures are supported. Suppose the CreateAsset function has the following signature:
type Asset struct {
	ID             string `json:"ID"`
	Color          string `json:"color"`
	Size           int    `json:"size"`
	Owner          string `json:"owner"`
	Appraisal *Appraisal `json:"appraisal"`
}

type Appraisal struct {
	AppraisedValue int  `json:"appraisedValue"`
	Inspected      bool `json:"inspected"`
}

// CreateAsset issues a new asset to the world state with given details.
func (s *SmartContract) CreateAsset(ctx contractapi.TransactionContextInterface, id string, color string, size int, owner string, appraisal Appraisal) error {
  // implementation...
}

Note that the appraisal parameter is a complex type, the transaction input data can be specified as follows:

POST http://localhost:3000/transactions?fly-sync=true&fly-signer=user001&fly-channel=default-channel&fly-chaincode=asset_transfer
{
    "headers": {
        "type": "SendTransaction",
        "payloadSchema": {
            "type": "array",
            "prefixItems": [{
                "name": "id", "type": "string"
            }, {
                "name": "color", "type": "string"
            }, {
                "name": "size", "type": "integer"
            }, {
                "name": "owner", "type": "string"
            }, {
                "name": "appraisal", "type": "object",
                "properties": {
                    "appraisedValue": {
                        "type": "integer"
                    },
                    "inspected": {
                        "type": "boolean"
                    }
                }
            }]
        }
    },
    "func": "CreateAsset",
    "args": {
        "owner": "Tom",
        "appraisal": {
            "appraisedValue": 123000,
            "inspected": true
        },
        "size": 10,
        "id": "asset205",
        "color": "red"
    }
}

JSON Data Support in Events

If a chaincode publishes events with string or JSON data, fabconnect can be instructed to decode them from the byte array before sending the event to the listening client application. The decoding instructions can be provided during subscription.

For example, the following chaincode publishes an event containing a JSON structure in the payload:

	asset := Asset{
		ID:             id,
		Color:          color,
		Size:           size,
		Owner:          owner,
		Appraisal: &Appraisal{
			AppraisedValue: appraisal.AppraisedValue,
			Inspected: appraisal.inspected,
		},
	}
	assetJSON, _ := json.Marshal(asset)
	ctx.GetStub().SetEvent("AssetCreated", assetJSON)

An event subscription can be created as follows which contains instructions to decode the payload bytes:

{
  "stream": "es-31e85b01-6440-4cc3-63e9-2aafc0d06466",
  "channel": "default-channel",
  "name": "sub-1",
  "signer": "user001",
  "fromBlock": "100",
  "filter": {
    "chaincodeId": "assettransfercomplex"
  },
  "payloadType": "stringifiedJSON"
}

Notice the payloadType property, which instructs fabconnect to decode the payload bytes into a JSON structure. As a result the client will receive the event JSON as follows:

[
  {
    "chaincodeId": "assettransfercomplex",
    "blockNumber": 151,
    "transactionId": "8692254ea13e9f5cb021b613e722ce4610daa5c4529e1a9161497308b0278ca0",
    "eventName": "AssetCreated",
    "payload": {
      "ID": "asset204",
      "appraisal": {
        "appraisedValue": 123000,
        "inspected": true
      },
      "color": "red",
      "owner": "Tom",
      "size": 10
    },
    "subId": "sb-6859f687-61dd-44e8-6e8b-ddcf3b95b840"
  }
]

Besides stringifiedJSON, string is also supported as the payload type which represents UTF-8 encoded strings.

Fixes Needed for multiple subscriptions under the same event stream

The current fabric-sdk-go uses an internal cache for event services, which builds keys only using the channel ID. This means if there are multiple subscriptions targeting the same channel, but specify different fromBlock parameters, only the first instance will be effective. All subsequent subscriptions will share the same event service, rendering their own fromBlock configuration ineffective.

A fix has been provided for this in the forked repository https://github.com/kaleido-io/fabric-sdk-go.

Follow these simple steps to integrate this fix (until it's contributed back to the official repo):

  • clone the repository https://github.com/kaleido-io/fabric-sdk-go and place it peer to the firefly-fabconnect folder:
    workspace-root
       \_ fabric-sdk-go
       \_ firefly-fabconnect
    
  • checkout branch eventservice-cache-key
  • configure go to use it instead of the official package:
    go mod edit -replace=github.com/hyperledger/fabric-sdk-go=../fabric-sdk-go
    
  • rebuild with make

License

This project is licensed under the Apache 2 License - see the LICENSE file for details.

firefly-fabconnect's People

Contributors

janrockdev avatar jimthematrix avatar nguyer avatar peterbroadhurst avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.