fabric1.4.2 migrated from kafka to raft (command version)

See in detail the principles on which implementation depends. https://blog.csdn.net/qq_42750537/article/details/96428765

This article is used to record specific orders and problems encountered.

Constructing RatfMetadata

Get available configurations

1) Creating a raft-based network (v1.4.2) requires the same certificate file as the existing kafka network

2) Execute the Create Channel -> Execute Transaction commands in the new raft network

3) Execute the following commands to pull the configuration information of the system channel

Switching environment
export CORE_PEER_LOCALMSPID="OrdererMSP"

export CORE_PEER_TLS_ROOTCERT_FILE=/opt/gopath/src/github.com/hyperledger/fabric/peer/crypto-config/ordererOrganizations/example.com/orderers/orderer1.example.com/msp/tlscacerts/tlsca.example.com-cert.pem

export CORE_PEER_MSPCONFIGPATH=/opt/gopath/src/github.com/hyperledger/fabric/peer/crypto-config/ordererOrganizations/example.com/users/Admin@example.com/msp

export ORDERER_CA=/opt/gopath/src/github.com/hyperledger/fabric/peer/addorg/crypto-config/ordererOrganizations/example.com/orderers/orderer1.example.com/msp/tlscacerts/tlsca.example.com-cert.pem

export CH_NAME=testchainid

export CONFIGTXLATOR_URL=http://127.0.0.1:7059

configtxlator start > log.log 2>&1 &

echo $CH_NAME

//Generating Readable Configuration Block Files
peer channel fetch config config_block.pb -o orderer1.example.com:7050 -c $CH_NAME --tls --cafile $ORDERER_CA

curl -X POST --data-binary @config_block.pb "$CONFIGTXLATOR_URL/protolator/decode/common.Block" | jq . > config_block.json

jq .data.data[0].payload.data.config config_block.json> config.json

Building Metadata

1) Open config.json. Search Consensus Type.

2) Remove the Consensus Type section and modify the status to STATE_MAINTENANCE. Save as file raft.json

Enter Maintenance Status

Converting the configuration of each channel in the kafak-fabric network

Execute the following commands

1.Switching environment
2.Generating Readable Configuration Block Files
3.Modify configuration block information
jq ".channel_group.groups.Orderer.values.ConsensusType.value.state = \"STATE_MAINTENANCE\"" config.json  > modified_config.json
4.Combine the modified information with the original configuration information
curl -X POST --data-binary @config.json "$CONFIGTXLATOR_URL/protolator/encode/common.Config" > config.pb

curl -X POST --data-binary @modified_config.json "$CONFIGTXLATOR_URL/protolator/encode/common.Config" > modified_config.pb

curl -X POST -F channel=$CH_NAME -F "original=@config.pb" -F "updated=@modified_config.pb" "${CONFIGTXLATOR_URL}/configtxlator/compute/update-from-configs" > config_update.pb

curl -X POST --data-binary @config_update.pb "$CONFIGTXLATOR_URL/protolator/decode/common.ConfigUpdate" | jq . > config_update.json

echo '{"payload":{"header":{"channel_header":{"channel_id":"'$CH_NAME'", "type":2}},"data":{"config_update":'$(cat config_update.json)'}}}' | jq . > config_update_in_envelope.json

curl -X POST --data-binary @config_update_in_envelope.json "$CONFIGTXLATOR_URL/protolator/encode/common.Envelope" > config_update_in_envelope.pb
5.Get more than half of the organizational signatures (only signature commands are written here)
export CORE_PEER_LOCALMSPID=OrdererMSP
export CORE_PEER_MSPCONFIGPATH=/opt/gopath/src/github.com/hyperledger/fabric/peer/crypto-config/ordererOrganizations/example.com/users/Admin\@example.com/msp
export CORE_PEER_ADDRESS=peer0.org1:7051
export CORE_PEER_TLS_ENABLED=true
export CORE_PEER_TLS_ROOTCERT_FILE=/opt/gopath/src/github.com/hyperledger/fabric/peer/crypto-config/ordererOrganizations/example.com/orderers/orderer1.example.com/tls/ca.crt
export CORE_PEER_TLS_KEY_FILE=/opt/gopath/src/github.com/hyperledger/fabric/peer/crypto-config/ordererOrganizations/ecample.com/orderers/orderer1.example.com/tls/server.key
export CORE_PEER_TLS_CERT_FILE=/opt/gopath/src/github.com/hyperledger/fabric/peer/addorg/crypto-config/ordererOrganizations/example.com/orderers/orderer1.example.com/tls/server.crt
peer channel signconfigtx -f config_update_in_envelope.pb
6.Update configuration
peer channel update -f config_update_in_envelope.pb -c $CH_NAME -o orderer1.example.com:7050 --tls true --cafile $ORDERER_CA

pkill configtxlator

Switch export CH_NAME = other user channels to perform the above actions in turn

Back up files and shut down the server

1. Optional, mainly rollback when migration fails in production environments. Test environments are not executable

Switch to Raft in Maintenance Mode

Create a new scripts folder and import raft.json that you prepared earlier.

Execute the following commands on each channel in turn

1.Switching environment
2.Generating Readable Configuration Block Files
3.Modify configuration block information
jq -s '.[0]*{"channel_group":{"groups":{"Orderer": {"values": {"ConsensusType": .[1]}}}}}' config.json ./scripts/raft.json > modified_config.json
4.Combine the modified information with the original configuration information
5.Get more than half of the organizational signatures (only signature commands are written here)
6.Update configuration

Switch export CH_NAME = other user channels to perform the above actions in turn

Restart and validate leaders

Stop all sorting service nodes, stop all Kafka agents and Zookeepers, and then restart only sorting service nodes.

Start successfully to view the orderer log

Successful elections

Switch to normal mode

But at this time, the normal transaction can not be completed, and the fast configuration state needs to be switched to NORMAL.

Execute the following commands on each channel in turn

1.Switching environment
2.Generating Readable Configuration Block Files
3.Modify configuration block information
jq ".channel_group.groups.Orderer.values.ConsensusType.value.state = \"STATE_NORMAL\"" config.json  > modified_config.json
4.Combine the modified information with the original configuration information
5.Get more than half of the organizational signatures (only signature commands are written here)
6.Update configuration

Switch export CH_NAME = other user channels to perform the above actions in turn

After completing this process, the sorting service can now accept all transactions across all channels. If you stop peers and applications as recommended, you can restart them now.

Problem Point Record

1. Sort Node Restart Error Reporting: no such hosts:

Reason for the problem: The mapping of the original fabric network orderer image does not contain the mapping of orderer nodes

Solution: Enter the orderer container and add the orderer node mapping

2: Sort node restart error reporting, tls connection failed

Reason: The tls-related configuration of the original yaml file is identified by kafka, and raft is not recognized. New raft identifiable configuration is required

The solution is to provide export updates to the orderer container for the corresponding configuration. Then restart the container.