We're stuck configuring a fabric network based on 3 orgs with 1 peer each and 2 kafka-based orderers. For kafka ordering we use 4 kafka nodes with 3 zookeepers. It's deployed on some AWS ec2 instances as follows:
1: Org1
2: Org2
3: Org3
4: orderer0, orderer1, kafka0, kafka1, kafka2, kafka3, zookeeper0, zookeeper1, zookeeper2
The whole of the ordering nodes plus kafka cluster is located in the same machine for connectivity reasons (read somewhere they must be in the same machine to avoid these problems)
During our test, we try taking down the first orderer (orderer0) for redundancy testing with docker stop. We expected the network to continue working through orderer1, but instead it dies and stops working.
Looking at the peer's console, I can see some errors.
Could not connect to any of the endpoints: [orderer0.example.com:7050, orderer1.example.com:8050]
Find attached the content of the files related to the configuration of the system.
Orderer + kafka + zk
#
# Copyright IBM Corp. All Rights Reserved.
#
# SPDX-License-Identifier: Apache-2.0
#
version: '2'
services:
zookeeper0.example.com:
container_name: zookeeper0.example.com
extends:
file: docker-compose-base.yaml
service: zookeeper0.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
zookeeper1.example.com:
container_name: zookeeper1.example.com
extends:
file: docker-compose-base.yaml
service: zookeeper1.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
zookeeper2.example.com:
container_name: zookeeper2.example.com
extends:
file: docker-compose-base.yaml
service: zookeeper2.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
kafka0.example.com:
container_name: kafka0.example.com
extends:
file: docker-compose-base.yaml
service: kafka0.example.com
depends_on:
- zookeeper0.example.com
- zookeeper1.example.com
- zookeeper2.example.com
- orderer0.example.com
- orderer1.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
kafka1.example.com:
container_name: kafka1.example.com
extends:
file: docker-compose-base.yaml
service: kafka1.example.com
depends_on:
- zookeeper0.example.com
- zookeeper1.example.com
- zookeeper2.example.com
- orderer0.example.com
- orderer1.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
kafka2.example.com:
container_name: kafka2.example.com
extends:
file: docker-compose-base.yaml
service: kafka2.example.com
depends_on:
- zookeeper0.example.com
- zookeeper1.example.com
- zookeeper2.example.com
- orderer0.example.com
- orderer1.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
kafka3.example.com:
container_name: kafka3.example.com
extends:
file: docker-compose-base.yaml
service: kafka3.example.com
depends_on:
- zookeeper0.example.com
- zookeeper1.example.com
- zookeeper2.example.com
- orderer0.example.com
- orderer1.example.com
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
orderer0.example.com:
container_name: orderer0.example.com
image: hyperledger/fabric-orderer:x86_64-1.1.0
environment:
- ORDERER_GENERAL_LOGLEVEL=debug
- ORDERER_GENERAL_LISTENADDRESS=0.0.0.0
- ORDERER_LISTEN_PORT=7050
- ORDERER_GENERAL_GENESISMETHOD=file
- ORDERER_GENERAL_GENESISFILE=/etc/hyperledger/configtx/genesis.block
- ORDERER_GENERAL_LOCALMSPID=OrdererMSP
- ORDERER_GENERAL_LOCALMSPDIR=/etc/hyperledger/crypto/orderer/msp
- ORDERER_GENERAL_TLS_ENABLED=true
- ORDERER_GENERAL_TLS_PRIVATEKEY=/etc/hyperledger/crypto/orderer/tls/server.key
- ORDERER_GENERAL_TLS_CERTIFICATE=/etc/hyperledger/crypto/orderer/tls/server.crt
- ORDERER_GENERAL_TLS_ROOTCAS=[/etc/hyperledger/crypto/orderer/tls/ca.crt, /etc/hyperledger/crypto/peerOrg1/tls/ca.crt, /etc/hyperledger/crypto/peerOrg2/tls/ca.crt, /etc/hyperledger/crypto/peerOrg3/tls/ca.crt]
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
working_dir: /opt/gopath/src/github.com/hyperledger/fabric/orderers
command: orderer
ports:
- 7050:7050
volumes:
- ./channel:/etc/hyperledger/configtx
- ./channel/crypto-config/ordererOrganizations/example.com/orderers/orderer0.example.com/:/etc/hyperledger/crypto/orderer
- ./channel/crypto-config/peerOrganizations/org1.example.com/peers/peer0.org1.example.com/:/etc/hyperledger/crypto/peerOrg1
- ./channel/crypto-config/peerOrganizations/org2.example.com/peers/peer0.org2.example.com/:/etc/hyperledger/crypto/peerOrg2
- ./channel/crypto-config/peerOrganizations/org3.example.com/peers/peer0.org3.example.com/:/etc/hyperledger/crypto/peerOrg3
depends_on:
- kafka0.example.com
- kafka1.example.com
- kafka2.example.com
- kafka3.example.com
orderer1.example.com:
container_name: orderer1.example.com
image: hyperledger/fabric-orderer:x86_64-1.1.0
environment:
- ORDERER_GEN ERAL_LOGLEVEL=debug
- ORDERER_GENERAL_LISTENADDRESS=0.0.0.0
- ORDERER_LISTEN_PORT=8050
- ORDERER_GENERAL_GENESISMETHOD=file
- ORDERER_GENERAL_GENESISFILE=/etc/hyperledger/configtx/genesis.block
- ORDERER_GENERAL_LOCALMSPID=OrdererMSP
- ORDERER_GENERAL_LOCALMSPDIR=/etc/hyperledger/crypto/orderer/msp
- ORDERER_GENERAL_TLS_ENABLED=true
- ORDERER_GENERAL_TLS_PRIVATEKEY=/etc/hyperledger/crypto/orderer/tls/server.key
- ORDERER_GENERAL_TLS_CERTIFICATE=/etc/hyperledger/crypto/orderer/tls/server.crt
- ORDERER_GENERAL_TLS_ROOTCAS=[/etc/hyperledger/crypto/orderer/tls/ca.crt, /etc/hyperledger/crypto/peerOrg1/tls/ca.crt, /etc/hyperledger/crypto/peerOrg2/tls/ca.crt, /etc/hyperledger/crypto/peerOrg3/tls/ca.crt]
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
working_dir: /opt/gopath/src/github.com/hyperledger/fabric/orderers
command: orderer
ports:
- 8050:7050
volumes:
- ./channel:/etc/hyperledger/configtx
- ./channel/crypto-config/ordererOrganizations/example.com/orderers/orderer1.example.com/:/etc/hyperledger/crypto/orderer
- ./channel/crypto-config/peerOrganizations/org1.example.com/peers/peer0.org1.example.com/:/etc/hyperledger/crypto/peerOrg1
- ./channel/crypto-config/peerOrganizations/org2.example.com/peers/peer0.org2.example.com/:/etc/hyperledger/crypto/peerOrg2
- ./channel/crypto-config/peerOrganizations/org3.example.com/peers/peer0.org3.example.com/:/etc/hyperledger/crypto/peerOrg3
depends_on:
- kafka0.example.com
- kafka1.example.com
- kafka2.example.com
- kafka3.example.com
Peer and Ca from Org2
#
# Copyright IBM Corp. All Rights Reserved.
#
# SPDX-License-Identifier: Apache-2.0
#
version: '2'
services:
ca.org2.example.com:
image: hyperledger/fabric-ca:x86_64-1.1.0
environment:
- FABRIC_CA_HOME=/etc/hyperledger/fabric-ca-server
- FABRIC_CA_SERVER_CA_CERTFILE=/etc/hyperledger/fabric-ca-server-config/ca.org2.example.com-cert.pem
- FABRIC_CA_SERVER_CA_KEYFILE=/etc/hyperledger/fabric-ca-server-config/efa7d0819b7083f6c06eb34da414acbcde79f607b9ce26fb04dee60cf79a389a_sk
- FABRIC_CA_SERVER_TLS_ENABLED=true
- FABRIC_CA_SERVER_TLS_CERTFILE=/etc/hyperledger/fabric-ca-server-config/ca.org2.example.com-cert.pem
- FABRIC_CA_SERVER_TLS_KEYFILE=/etc/hyperledger/fabric-ca-server-config/efa7d0819b7083f6c06eb34da414acbcde79f607b9ce26fb04dee60cf79a389a_sk
ports:
- "8054:7054"
command: sh -c 'fabric-ca-server start -b admin:adminpw -d'
volumes:
- ./channel/crypto-config/peerOrganizations/org2.example.com/ca/:/etc/hyperledger/fabric-ca-server-config
container_name: ca_peerOrg2
peer0.org2.example.com:
container_name: peer0.org2.example.com
extends:
file: base.yaml
service: peer-base
environment:
- CORE_PEER_ID=peer0.org2.example.com
- CORE_PEER_LOCALMSPID=Org2MSP
- CORE_PEER_ADDRESS=peer0.org2.example.com:7051
ports:
- 8051:7051
- 8053:7053
volumes:
- ./channel/crypto-config/peerOrganizations/org2.example.com/peers/peer0.org2.example.com/:/etc/hyperledger/crypto/peer
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
extra_hosts:
- "orderer0.example.com:xxx.xxx.xxx.xxx"
- "orderer1.example.com:xxx.xxx.xxx.xxx"
- "kafka0.example.com:xxx.xxx.xxx.xxx"
- "kafka1.example.com:xxx.xxx.xxx.xxx"
- "kafka2.example.com:xxx.xxx.xxx.xxx"
- "kafka3.example.com:xxx.xxx.xxx.xxx"
Orderer base
# Copyright IBM Corp. All Rights Reserved.
#
# SPDX-License-Identifier: Apache-2.0
#
version: '2'
services:
orderer-base:
image: hyperledger/fabric-orderer:$IMAGE_TAG
environment:
- ORDERER_GENERAL_LOGLEVEL=error
- ORDERER_GENERAL_LISTENADDRESS=0.0.0.0
- ORDERER_GENERAL_GENESISMETHOD=file
- ORDERER_GENERAL_GENESISFILE=/var/hyperledger/orderer/orderer.genesis.block
- ORDERER_GENERAL_LOCALMSPID=OrdererMSP
- ORDERER_GENERAL_LOCALMSPDIR=/var/hyperledger/orderer/msp
# enabled TLS
- ORDERER_GENERAL_TLS_ENABLED=true
- ORDERER_GENERAL_TLS_PRIVATEKEY=/var/hyperledger/orderer/tls/server.key
- ORDERER_GENERAL_TLS_CERTIFICATE=/var/hyperledger/orderer/tls/server.crt
- ORDERER_GENERAL_TLS_ROOTCAS=[/var/hyperledger/orderer/tls/ca.crt]
# kafka
- CONFIGTX_ORDERER_ORDERERTYPE=kafka
- CONFIGTX_ORDERER_KAFKA_BROKERS=[kafka0.example.com,kafka1.example.com,kafka2.example.com,kafka3.example.com]
- ORDERER_KAFKA_RETRY_SHORTINTERVAL=1s
- ORDERER_KAFKA_RETRY_SHORTTOTAL=30s
- ORDERER_KAFKA_VERBOSE=true
working_dir: /opt/gopath/src/github.com/hyperledger/fabric
command: orderer
Kafka base
# Copyright IBM Corp. All Rights Reserved.
#
# SPDX-License-Identifier: Apache-2.0
#
version: '2'
services:
zookeeper:
image: hyperledger/fabric-zookeeper
environment:
- ZOO_SERVERS=server.1=zookeeper0.example.com:2888:3888 server.2=zookeeper1.example.com:2888:3888 server.3=zookeeper2.example.com:2888:3888
restart: always
kafka:
image: hyperledger/fabric-kafka
restart: always
environment:
- KAFKA_MESSAGE_MAX_BYTES=103809024 # 99 * 1024 * 1024 B
- KAFKA_REPLICA_FETCH_MAX_BYTES=103809024 # 99 * 1024 * 1024 B
- KAFKA_UNCLEAN_LEADER_ELECTION_ENABLE=false
- KAFKA_MIN_INSYNC_REPLICAS=2
- KAFKA_DEFAULT_REPLICATION_FACTOR=3
- KAFKA_ZOOKEEPER_CONNECT=zookeeper0.example.com:2181,zookeeper1.example.com:2181,zookeeper2.example.com:2181
logging:
driver: "json-file"
options:
max-size: "1m"
max-file: "3"
configtx.yaml
Organizations:
- &OrdererOrg
Name: OrdererMSP
ID: OrdererMSP
MSPDir: crypto-config/ordererOrganizations/example.com/msp
- &Org1
Name: Org1MSP
ID: Org1MSP
MSPDir: crypto-config/peerOrganizations/org1.example.com/msp
AnchorPeers:
- Host: peer0.org1.example.com
Port: 7051
- &Org2
Name: Org2MSP
ID: Org2MSP
MSPDir: crypto-config/peerOrganizations/org2.example.com/msp
AnchorPeers:
- Host: peer0.org2.example.com
Port: 7051
- &Org3
Name: Org3MSP
ID: Org3MSP
MSPDir: crypto-config/peerOrganizations/org3.example.com/msp
AnchorPeers:
- Host: peer0.org3.example.com
Port: 7051
################################################################################
Orderer: &OrdererDefaults
OrdererType: kafka
Addresses:
- orderer0.example.com:7050
- orderer1.example.com:7050
BatchTimeout: 2s
BatchSize:
MaxMessageCount: 10
AbsoluteMaxBytes: 98 MB
PreferredMaxBytes: 512 KB
Kafka:
Brokers:
- kafka0.example.com:9092
- kafka1.example.com:9092
- kafka2.example.com:9092
- kafka3.example.com:9092
Organizations:
################################################################################
Application: &ApplicationDefaults
Organizations:
################################################################################
Profiles:
ThreeOrgsOrdererGenesis:
Orderer:
<<: *OrdererDefaults
Organizations:
- *OrdererOrg
Consortiums:
SampleConsortium:
Organizations:
- *Org1
- *Org2
- *Org3
ThreeOrgsChannel:
Consortium: SampleConsortium
Application:
<<: *ApplicationDefaults
Organizations:
- *Org1
- *Org2
- *Org3
May it be a configuration error? Connection problems are almost discarded because running the same network on a local machine gives the same result.
Thanks in advance Regards
Finally got it running smooth. Turns out the problem wasn't in docker-compose files, but in the version of fabric sdk for the web service. I was using fabric-client and fabric-ca-client both on version 1.1, and this was missing until 1.2. (More info https://jira.hyperledger.org/browse/FABN-90)
Just to clarify, I was able to see transactions happening on both orderers because of the interconnection between them, but I was only attacking the first one. When that orderer went down, network would go dark.
I understood the way fabric deals with orderers, it points to the first orderer of the list, if it is down or unreachable, moves it to the bottom of the list and targets the next one. This is what's happening since 1.2, in older versions you have to code your own error controller so that it changes to the next orderer.
I'm not sure but it could be because of different network layer. Since it's a different compose file , docker do create different network layer for each composer.
Also, I don't see network mentioned in the yaml files.
Please check list of network layer using "docker network list".
来源：https://stackoverflow.com/questions/52606852/multiple-orderer-redundancy-in-kafka-based-network