In the following repository there are basic skeletons for the network creation.

https://github.com/mtnieto/hf-template

• Go 1.10 version

• Ubuntu 16.04

Execute the following commands to have the environment ready.

sudo apt install libtool libltdl-dev // with mac brew intall libtool
go get github.com/miekg/pkcs11
curl -sSL http:https://bit.ly/2ysbOFE | bash -s 1.3.0 // or other version

The following files will be modified to create the configuration desired.

In this file is established the definition of the organizations, peers for organization and the configuration of the orederer.

In this file is established the channel configurations, how many channel, what organizations are in a channel, etc.

It is necessary to change the hosts and the paths of the crypto materials according the configuration of the 2 files previous.

It is also necessary necessary to change the hosts and the paths of the crypto materials.

Modify the two following methods

1. replacePrivateKey: Change the paths of the cryptomaterials

2. generateChannelArtifacts: Modify the following command lines

$GOPATH/bin/configtxgen -profile SeriesOrdererGenesis -outputBlock ./channel-artifacts/genesis.block

$GOPATH/bin/configtxgen -profile serieschannel -outputCreateChannelTx ./channel-artifacts/serieschannel.tx -channelID "serieschannel"

It is necessary to be equal to the variables defined in the configtx.yaml file. Be careful with the channelID, uppercases are considered like ilegal characters.

Once the configuration is defined and all the files modified, to run the network execute the following command:

./fabricOps.sh start

NOTE: Now, you have an environment to deploy and to interact chaincodes. It's very important that the chaincode is deployed in the same docker network than the peers. To check this execute the following command

docker inspect "name_of_a_peer_container

With this information, the next step is going to the file /base/peer-base.yaml.

The name of the network must correspond to the following line value. If it is different, when a chaincode will be instanciate its docker container will be created and immediately will be killed.

- CORE_VM_DOCKER_HOSTCONFIG_NETWORKMODE=hf2orgs_default

With every change in the configuration it is necessary to execute the following command:

./fabricOps.sh clean

NOTE: You must not delete the command line in the script where the docker images are removed.

To interact with the network creating channels, joining peers and interacting with the chaincode, execute the following command:

./fabricOps.sh cli

Once we are in the docker container of the hyperledger cli tool we can execute the following commands:

./scripts/01-createchannel.sh //a channel is created
./scripts/02-joinHBO.sh // An organization joins in to the channel
./scripts/03-joinNetflix.sh //Another organization joins into the channel
./script/04-installHBO.sh // The chaincode is installed in the hbo peer
./script/06-instanciate.sh // The chaincode is instanciated in the network and a docker-container of the chaincode is created

Everytime we are going to distribute the network it is very important to generate first of all the certificates, and then distribute the code to the different machines.

With this goal, comment the following lines of the file fabricOps.sh.

 generateCerts
 generateChannelArtifacts
 replacePrivateKey
# pullDockerImages
# startNetwork

Now, the crypto materials were generated. The next step is to comment and to uncomment the following lines of the fabricOps.sh

# generateCerts
# generateChannelArtifacts
# replacePrivateKey
pullDockerImages
# startNetwork

It is necessary to modify the following files:

Append the following lines to the document with the configuration desired:

extra_hosts:
  - "orderer.myapp.com:34.240.181.209"
  - "peer0.netflix.myapp.com:34.240.181.209"
  - "peer0.hbo.myapp.com:34.248.224.23"

Append the same lines in the configuration of the cli container

We are going to use 2 machines. In the first machine will run the orderer, one peer and its ca, and the cli tool.

In the second machine will run the other peer and its ca.

The commands to run the distributed network are the following.

First machine:

docker-compose up -d cli

Second machine:

docker-compose up -d peer0.hbo.myapp.com

To check if the peers of the networks are seen between them it is necessary to execute the scritps. The scripts create a channel, join peers into the channel, install the chaincode in the peers, and instanciate the chaincode in one peer.

To run the scripts execute the following command:

./fabricOps.sh cli
./scripts/01-createchannel.sh //a channel is created
./scripts/02-joinHBO.sh // An organization joins in to the channel
./scripts/03-joinNetflix.sh //Another organization joins into the channel
./script/04-installHBO.sh // The chaincode is installed in the hbo peer
./script/06-instanciate.sh // The chaincode is instanciated in the network and a docker-container of the chaincode is created

If the execution is correct and if in the second machine a container is created, the network is running!!