作者：Tatsuya Naganawa 译者：TF编译组
多kube-master部署
3个Tungsten Fabric控制器节点：m3.xlarge（4 vcpu）-> c3.4xlarge（16 vcpu）（由于schema-transformer需要cpu资源进行acl计算，因此我需要添加资源）
100 kube-master, 800 workers: m3.medium
在下面这个链接内容中，tf-controller安装和first-containers.yaml是相同的
- https://github.com/tnaganawa/tungstenfabric-docs/blob/master/TungstenFabricPrimer.md#2-tungstenfabric-up-and-running
Ami也相同（ami-3185744e），但是内核版本通过yum -y update kernel（转换为映像，并用于启动实例）更新
/tmp/aaa.pem是ec2实例中指定的密钥对
附cni.yaml文件：
(在其中一个Tungsten Fabric控制器节点键入命令)
yum -y install epel-release
yum -y install parallel
aws ec2 describe-instances --query 'Reservations[*].Instances[*].PrivateIpAddress' --output text | tr '\t' '\n' > /tmp/all.txt
head -n 100 /tmp/all.txt > masters.txt
tail -n 800 /tmp/all.txt > workers.txt
ulimit -n 4096
cat all.txt | parallel -j1000 ssh -i /tmp/aaa.pem -o StrictHostKeyChecking=no centos@{} id
cat all.txt | parallel -j1000 ssh -i /tmp/aaa.pem centos@{} sudo sysctl -w net.bridge.bridge-nf-call-iptables=1
-
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' ssh -i /tmp/aaa.pem centos@{2} sudo kubeadm init --token aaaaaa.aaaabbbbccccdddd --ignore-preflight-errors=NumCPU --pod-network-cidr=10.32.{1}.0/24 --service-cidr=10.96.{1}.0/24 --service-dns-domain=cluster{1}.local
-
vi assign-kube-master.py
computenodes=8
with open ('masters.txt') as aaa:
with open ('workers.txt') as bbb:
for masternode in aaa.read().rstrip().split('\n'):
for i in range (computenodes):
tmp=bbb.readline().rstrip()
print ("{}\t{}".format(masternode, tmp))
-
cat masters.txt | parallel -j1000 ssh -i /tmp/aaa.pem centos@{} sudo cp /etc/kubernetes/admin.conf /tmp/admin.conf
cat masters.txt | parallel -j1000 ssh -i /tmp/aaa.pem centos@{} sudo chmod 644 /tmp/admin.conf
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' scp -i /tmp/aaa.pem centos@{2}:/tmp/admin.conf kubeconfig-{1}
-
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' kubectl --kubeconfig=kubeconfig-{1} get node
-
cat -n join.txt | parallel -j1000 -a - --colsep '\t' ssh -i /tmp/aaa.pem centos@{3} sudo kubeadm join {2}:6443 --token aaaaaa.aaaabbbbccccdddd --discovery-token-unsafe-skip-ca-verification
-
(modify controller-ip in cni-tungsten-fabric.yaml)
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' cp cni-tungsten-fabric.yaml cni-{1}.yaml
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' sed -i -e "s/k8s2/k8s{1}/" -e "s/10.32.2/10.32.{1}/" -e "s/10.64.2/10.64.{1}/" -e "s/10.96.2/10.96.{1}/" -e "s/172.31.x.x/{2}/" cni-{1}.yaml
-
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' kubectl --kubeconfig=kubeconfig-{1} apply -f cni-{1}.yaml
-
sed -i 's!kubectl!kubectl --kubeconfig=/etc/kubernetes/admin.conf!' set-label.sh
cat masters.txt | parallel -j1000 scp -i /tmp/aaa.pem set-label.sh centos@{}:/tmp
cat masters.txt | parallel -j1000 ssh -i /tmp/aaa.pem centos@{} sudo bash /tmp/set-label.sh
-
cat -n masters.txt | parallel -j1000 -a - --colsep '\t' kubectl --kubeconfig=kubeconfig-{1} create -f first-containers.yaml在OpenStack上嵌套安装Kubernetes
可以在all-in-one的openstack节点上尝试嵌套安装kubernetes。
在ansible-deployer安装了该节点之后，
-
https://github.com/tnaganawa/tungstenfabric-docs/blob/master/TungstenFabricPrimer.md#openstack-1
此外，需要手动创建为vRouter TCP/9091连接本地服务 - https://github.com/Juniper/contrail-kubernetes-docs/blob/master/install/kubernetes/nested-kubernetes.md#option-1-fabric-snat--link-local-preferred
此配置将创建DNAT/SNAT，例如从src: 10.0.1.3:xxxx, dst-ip: 10.1.1.11:9091到src: compute's vhost0 ip:xxxx dst-ip: 127.0.0.1:9091，因此openstack VM中的CNI可以与计算节点上的vrouter-agent直接通信，并为容器选择端口/ip信息。
IP地址可以来自子网，也可以来自子网外部。
在该节点上，将创建两个Centos7（或ubuntu bionic）节点，并将使用相同的程序（见下面链接）安装kubernetes集群，
- https://github.com/tnaganawa/tungstenfabric-docs/blob/master/TungstenFabricPrimer.md#kubeadm
当然yaml文件需要嵌套安装。
./resolve-manifest.sh contrail-nested-kubernetes.yaml > cni-tungsten-fabric.yaml
KUBEMANAGER_NESTED_MODE: "{{ KUBEMANAGER_NESTED_MODE }}" ## this needs to be "1"
KUBERNESTES_NESTED_VROUTER_VIP: {{ KUBERNESTES_NESTED_VROUTER_VIP }} ## this parameter needs to be the same IP with the one defined in link-local service (such as 10.1.1.11)如果coredns接收到ip，则说明嵌套安装正常。
vRouter ml2插件
我尝试了vRouter neutron插件的ml2功能。
- https://opendev.org/x/networking-opencontrail/
- https://www.youtube.com/watch?v=4MkkMRR9U2s
使用AWS上的三个CentOS7.5（4 cpu，16 GB内存，30 GB磁盘，ami: ami-3185744e）。
随附基于本文件的步骤。
openstack-controller: 172.31.15.248
tungsten-fabric-controller (vRouter): 172.31.10.212
nova-compute (ovs): 172.31.0.231
(命令在tungsten-fabric-controller上，使用centos用户（不是root用户）)
sudo yum -y remove PyYAML python-requests
sudo yum -y install git patch
sudo easy_install pip
sudo pip install PyYAML requests ansible==2.8.8
ssh-keygen
add id_rsa.pub to authorized_keys on all three nodes (centos user (not root))
git clone https://opendev.org/x/networking-opencontrail.git
cd networking-opencontrail
patch -p1 < ml2-vrouter.diff
cd playbooks
cp -i hosts.example hosts
cp -i group_vars/all.yml.example group_vars/all.yml
(ssh to all the nodes once, to update known_hosts)
ansible-playbook main.yml -i hosts
- devstack日志位于/opt/stack/logs/stack.sh.log中
- openstack进程日志写在/var/log/messages中
- 'systemctl list-unit-files | grep devstack'显示openstack进程的systemctl条目
(openstack控制器节点)
一旦devstack因mariadb登录错误而失败，请键入此命令进行修复。（对于openstack控制器的ip和fqdn，需要修改最后两行）
命令将由“centos”用户（不是root用户）键入.
mysqladmin -u root password admin
mysql -uroot -padmin -h127.0.0.1 -e 'GRANT ALL PRIVILEGES ON *.* TO '\''root'\''@'\''%'\'' identified by '\''admin'\'';'
mysql -uroot -padmin -h127.0.0.1 -e 'GRANT ALL PRIVILEGES ON *.* TO '\''root'\''@'\''172.31.15.248'\'' identified by '\''admin'\'';'
mysql -uroot -padmin -h127.0.0.1 -e 'GRANT ALL PRIVILEGES ON *.* TO '\''root'\''@'\''ip-172-31-15-248.ap-northeast-1.compute.internal'\'' identified by '\''admin'\'';'下附hosts、group_vars/all和patch的代码。（有些仅是错误修正，但有些会更改默认行为）
[centos@ip-172-31-10-212 playbooks]$ cat hosts
controller ansible_host=172.31.15.248 ansible_user=centos
# 此host应为计算主机组中的一个.
# 单独部署Tungsten Fabtic计算节点的playbook尚未准备好.
contrail_controller ansible_host=172.31.10.212 ansible_user=centos local_ip=172.31.10.212
[contrail]
contrail_controller
[openvswitch]
other_compute ansible_host=172.31.0.231 local_ip=172.31.0.231 ansible_user=centos
[compute:children]
contrail
openvswitch
[centos@ip-172-31-10-212 playbooks]$ cat group_vars/all.yml
---
# IP address for OpenConrail (e.g. 192.168.0.2)
contrail_ip: 172.31.10.212
# Gateway address for OpenConrail (e.g. 192.168.0.1)
contrail_gateway:
# Interface name for OpenConrail (e.g. eth0)
contrail_interface:
# IP address for OpenStack VM (e.g. 192.168.0.3)
openstack_ip: 172.31.15.248
# 在VM上使用的OpenStack分支.
openstack_branch: stable/queens
# 也可以选择使用其它插件版本（默认为OpenStack分支）
networking_plugin_version: master
# Tungsten Fabric docker image tag for contrail-ansible-deployer
contrail_version: master-latest
# 如果为true，则使用Tungsten Fabric驱动程序安装networking_bgp***插件
install_networking_bgp***_plugin: false
# 如果true，则与设备管理器（将启动）和vRouter集成
# 封装优先级将被设置为'VXLAN,MPLSoUDP,MPLSoGRE'.
dm_integration_enabled: false
# 带有DM集成拓扑的文件的可选路径。设置并启用DM集成后，topology.yaml文件将复制到此位置
dm_topology_file:
# 如果为true，则将为当前ansible用户创建的实例密码设置为instance_password的值
change_password: false
# instance_password: uberpass1
# 如果已设置，请使用此数据覆盖docker daemon /etc config文件
# docker_config:
[centos@ip-172-31-10-212 playbooks]$
[centos@ip-172-31-10-212 networking-opencontrail]$ cat ml2-vrouter.diff
diff --git a/playbooks/roles/contrail_node/tasks/main.yml b/playbooks/roles/contrail_node/tasks/main.yml
index ee29b05..272ee47 100644
--- a/playbooks/roles/contrail_node/tasks/main.yml
+++ b/playbooks/roles/contrail_node/tasks/main.yml
@@ -7,7 +7,6 @@
- epel-release
- gcc
- git
- - ansible-2.4.*
- yum-utils
- libffi-devel
state: present
@@ -61,20 +60,20 @@
chdir: ~/contrail-ansible-deployer/
executable: /bin/bash
-- name: Generate ssh key for provisioning other nodes
- openssh_keypair:
- path: ~/.ssh/id_rsa
- state: present
- register: contrail_deployer_ssh_key
-
-- name: Propagate generated key
- authorized_key:
- user: "{{ ansible_user }}"
- state: present
- key: "{{ contrail_deployer_ssh_key.public_key }}"
- delegate_to: "{{ item }}"
- with_items: "{{ groups.contrail }}"
- when: contrail_deployer_ssh_key.public_key
+#- name: Generate ssh key for provisioning other nodes
+# openssh_keypair:
+# path: ~/.ssh/id_rsa
+# state: present
+# register: contrail_deployer_ssh_key
+#
+#- name: Propagate generated key
+# authorized_key:
+# user: "{{ ansible_user }}"
+# state: present
+# key: "{{ contrail_deployer_ssh_key.public_key }}"
+# delegate_to: "{{ item }}"
+# with_items: "{{ groups.contrail }}"
+# when: contrail_deployer_ssh_key.public_key
- name: Provision Node before deploy contrail
shell: |
@@ -105,4 +104,4 @@
sleep: 5
host: "{{ contrail_ip }}"
port: 8082
- timeout: 300
\ No newline at end of file
+ timeout: 300
diff --git a/playbooks/roles/contrail_node/templates/instances.yaml.j2 b/playbooks/roles/contrail_node/templates/instances.yaml.j2
index e3617fd..81ea101 100644
--- a/playbooks/roles/contrail_node/templates/instances.yaml.j2
+++ b/playbooks/roles/contrail_node/templates/instances.yaml.j2
@@ -14,6 +14,7 @@ instances:
config_database:
config:
control:
+ analytics:
webui:
{% if "contrail_controller" in groups["contrail"] %}
vrouter:
diff --git a/playbooks/roles/docker/tasks/main.yml b/playbooks/roles/docker/tasks/main.yml
index 8d7971b..5ed9352 100644
--- a/playbooks/roles/docker/tasks/main.yml
+++ b/playbooks/roles/docker/tasks/main.yml
@@ -6,7 +6,6 @@
- epel-release
- gcc
- git
- - ansible-2.4.*
- yum-utils
- libffi-devel
state: present
@@ -62,4 +61,4 @@
- docker-py==1.10.6
- docker-compose==1.9.0
state: present
- extra_args: --user
\ No newline at end of file
+ extra_args: --user
diff --git a/playbooks/roles/node/tasks/main.yml b/playbooks/roles/node/tasks/main.yml
index 0fb1751..d9ab111 100644
--- a/playbooks/roles/node/tasks/main.yml
+++ b/playbooks/roles/node/tasks/main.yml
@@ -1,13 +1,21 @@
---
-- name: Update kernel
+- name: Install required utilities
become: yes
yum:
- name: kernel
- state: latest
- register: update_kernel
+ name:
+ - python3-devel
+ - libibverbs ## needed by openstack controller node
+ state: present
-- name: Reboot the machine
- become: yes
- reboot:
- when: update_kernel.changed
- register: reboot_machine
+#- name: Update kernel
+# become: yes
+# yum:
+# name: kernel
+# state: latest
+# register: update_kernel
+#
+#- name: Reboot the machine
+# become: yes
+# reboot:
+# when: update_kernel.changed
+# register: reboot_machine
diff --git a/playbooks/roles/restack_node/tasks/main.yml b/playbooks/roles/restack_node/tasks/main.yml
index a11e06e..f66d2ee 100644
--- a/playbooks/roles/restack_node/tasks/main.yml
+++ b/playbooks/roles/restack_node/tasks/main.yml
@@ -9,7 +9,7 @@
become: yes
pip:
name:
- - setuptools
+ - setuptools==43.0.0
- requests
state: forcereinstall
[centos@ip-172-31-10-212 networking-opencontrail]$完成安装大约需要50分钟。
尽管/home/centos/devstack/openrc可以用于“demo”用户登录，但是需要管理员访问权限来指定其网络类型（vRouter为空，ovs为“vxlan”），因此需要手动创建adminrc。
[centos@ip-172-31-15-248 ~]$ cat adminrc
export OS_PROJECT_DOMAIN_ID=default
export OS_REGION_NAME=RegionOne
export OS_USER_DOMAIN_ID=default
export OS_PROJECT_NAME=admin
export OS_IDENTITY_API_VERSION=3
export OS_PASSWORD=admin
export OS_AUTH_TYPE=password
export OS_AUTH_URL=http://172.31.15.248/identity ## this needs to be modified
export OS_USERNAME=admin
export OS_TENANT_NAME=admin
export OS_VOLUME_API_VERSION=2
[centos@ip-172-31-15-248 ~]$
openstack network create testvn
openstack subnet create --subnet-range 192.168.100.0/24 --network testvn subnet1
openstack network create --provider-network-type vxlan testvn-ovs
openstack subnet create --subnet-range 192.168.110.0/24 --network testvn-ovs subnet1-ovs
- 创建了两个虚拟网络
[centos@ip-172-31-15-248 ~]$ openstack network list
+--------------------------------------+------------+--------------------------------------+
| ID | Name | Subnets |
+--------------------------------------+------------+--------------------------------------+
| d4e08516-71fc-401b-94fb-f52271c28dc9 | testvn-ovs | 991417ab-7da5-44ed-b686-8a14abbe46bb |
| e872b73e-100e-4ab0-9c53-770e129227e8 | testvn | 27d828eb-ada4-4113-a6f8-df7dde2c43a4 |
+--------------------------------------+------------+--------------------------------------+
[centos@ip-172-31-15-248 ~]$
- testvn's provider:network_type为空
[centos@ip-172-31-15-248 ~]$ openstack network show testvn
+---------------------------+--------------------------------------+
| Field | Value |
+---------------------------+--------------------------------------+
| admin_state_up | UP |
| availability_zone_hints | |
| availability_zones | nova |
| created_at | 2020-01-18T16:14:42Z |
| description | |
| dns_domain | None |
| id | e872b73e-100e-4ab0-9c53-770e129227e8 |
| ipv4_address_scope | None |
| ipv6_address_scope | None |
| is_default | None |
| is_vlan_transparent | None |
| mtu | 1500 |
| name | testvn |
| port_security_enabled | True |
| project_id | 84a573dbfadb4a198ec988e36c4f66f6 |
| provider:network_type | local |
| provider:physical_network | None |
| provider:segmentation_id | None |
| qos_policy_id | None |
| revision_number | 3 |
| router:external | Internal |
| segments | None |
| shared | False |
| status | ACTIVE |
| subnets | 27d828eb-ada4-4113-a6f8-df7dde2c43a4 |
| tags | |
| updated_at | 2020-01-18T16:14:44Z |
+---------------------------+--------------------------------------+
[centos@ip-172-31-15-248 ~]$
- 它是在Tungsten Fabric的数据库中创建的
(venv) [root@ip-172-31-10-212 ~]# contrail-api-cli --host 172.31.10.212 ls -l virtual-network
virtual-network/e872b73e-100e-4ab0-9c53-770e129227e8 default-domain:admin:testvn
virtual-network/5a88a460-b049-4114-a3ef-d7939853cb13 default-domain:default-project:dci-network
virtual-network/f61d52b0-6577-42e0-a61f-7f1834a2f45e default-domain:default-project:__link_local__
virtual-network/46b5d74a-24d3-47dd-bc82-c18f6bc706d7 default-domain:default-project:default-virtual-network
virtual-network/52925e2d-8c5d-4573-9317-2c346fb9edf0 default-domain:default-project:ip-fabric
virtual-network/2b0469cf-921f-4369-93a7-2d73350c82e7 default-domain:default-project:_internal_vn_ipv6_link_local
(venv) [root@ip-172-31-10-212 ~]#
- 另一方面，testvn-ovs's provider:network_type是vxlan，并且segmentation ID，mtu都是自动指定的
[centos@ip-172-31-15-248 ~]$ openstack network show testvn
+---------------------------+--------------------------------------+
| Field | Value |
+---------------------------+--------------------------------------+
| admin_state_up | UP |
| availability_zone_hints | |
| availability_zones | nova |
| created_at | 2020-01-18T16:14:42Z |
| description | |
| dns_domain | None |
| id | e872b73e-100e-4ab0-9c53-770e129227e8 |
| ipv4_address_scope | None |
| ipv6_address_scope | None |
| is_default | None |
| is_vlan_transparent | None |
| mtu | 1500 |
| name | testvn |
| port_security_enabled | True |
| project_id | 84a573dbfadb4a198ec988e36c4f66f6 |
| provider:network_type | local |
| provider:physical_network | None |
| provider:segmentation_id | None |
| qos_policy_id | None |
| revision_number | 3 |
| router:external | Internal |
| segments | None |
| shared | False |
| status | ACTIVE |
| subnets | 27d828eb-ada4-4113-a6f8-df7dde2c43a4 |
| tags | |
| updated_at | 2020-01-18T16:14:44Z |
+---------------------------+--------------------------------------+
[centos@ip-172-31-15-248 ~]$ openstack network show testvn-ovs
+---------------------------+--------------------------------------+
| Field | Value |
+---------------------------+--------------------------------------+
| admin_state_up | UP |
| availability_zone_hints | |
| availability_zones | nova |
| created_at | 2020-01-18T16:14:47Z |
| description | |
| dns_domain | None |
| id | d4e08516-71fc-401b-94fb-f52271c28dc9 |
| ipv4_address_scope | None |
| ipv6_address_scope | None |
| is_default | None |
| is_vlan_transparent | None |
| mtu | 1450 |
| name | testvn-ovs |
| port_security_enabled | True |
| project_id | 84a573dbfadb4a198ec988e36c4f66f6 |
| provider:network_type | vxlan |
| provider:physical_network | None |
| provider:segmentation_id | 50 |
| qos_policy_id | None |
| revision_number | 3 |
| router:external | Internal |
| segments | None |
| shared | False |
| status | ACTIVE |
| subnets | 991417ab-7da5-44ed-b686-8a14abbe46bb |
| tags | |
| updated_at | 2020-01-18T16:14:49Z |
+---------------------------+--------------------------------------+
[centos@ip-172-31-15-248 ~]$CentOS 8安装过程
centos8.2
ansible-deployer is used
only python3 is used (no python2)
- 需要ansible 2.8.x
1 x for tf-controller and kube-master, 1 vRouter
(all nodes)
yum install python3 chrony
alternatives --set python /usr/bin/python3
(vRouter nodes)
yum install network-scripts
- 这是必需的，因为vRouter当前不支持NetworkManager
(ansible node)
sudo yum -y install git
sudo pip3 install PyYAML requests ansible\
cirros-deployment-86885fbf85-tjkwn 1/1 Running 0 13s 10.47.255.249 ip-172-31-2-120.ap-northeast-1.compute.internal
[root@ip-172-31-7-20 ~]#
[root@ip-172-31-7-20 ~]#
[root@ip-172-31-7-20 ~]# kubectl exec -it cirros-deployment-86885fbf85-7z78k sh
/ # ip -o a
1: lo inet 127.0.0.1/8 scope host lo\ valid_lft forever preferred_lft forever
17: eth0 inet 10.47.255.250/12 scope global eth0\ valid_lft forever preferred_lft forever
/ # ping 10.47.255.249
PING 10.47.255.249 (10.47.255.249): 56 data bytes
64 bytes from 10.47.255.249: seq=0 ttl=63 time=0.657 ms
64 bytes from 10.47.255.249: seq=1 ttl=63 time=0.073 ms
^C
--- 10.47.255.249 ping statistics ---
2 packets transmitted, 2 packets received, 0% packet loss
round-trip min/avg/max = 0.073/0.365/0.657 ms
/ #
- 为了使chrony在安装路由器后正常工作，可能需要重新启动chrony服务器
[root@ip-172-31-4-206 ~]# chronyc -n sources
210 Number of sources = 5
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
^? 169.254.169.123 3 4 0 906 -8687ns[ -12us] +/- 428us
^? 129.250.35.250 2 7 0 1002 +429us[ +428us] +/- 73ms
^? 167.179.96.146 2 7 0 937 +665us[ +662us] +/- 2859us
^? 194.0.5.123 2 6 0 1129 +477us[ +473us] +/- 44ms
^? 103.202.216.35 3 6 0 933 +9662ns[+6618ns] +/- 145ms
[root@ip-172-31-4-206 ~]#
[root@ip-172-31-4-206 ~]#
[root@ip-172-31-4-206 ~]# service chronyd status
Redirecting to /bin/systemctl status chronyd.service
· chronyd.service - NTP client/server
Loaded: loaded (/usr/lib/systemd/system/chronyd.service; enabled; vendor preset: enabled)
Active: active (running) since Sun 2020-06-28 16:00:34 UTC; 33min ago
Docs: man:chronyd(8)
man:chrony.conf(5)
Main PID: 727 (chronyd)
Tasks: 1 (limit: 49683)
Memory: 2.1M
CGroup: /system.slice/chronyd.service
└─727 /usr/sbin/chronyd
Jun 28 16:00:33 localhost.localdomain chronyd[727]: Using right/UTC timezone to obtain leap second data
Jun 28 16:00:34 localhost.localdomain systemd[1]: Started NTP client/server.
Jun 28 16:00:42 localhost.localdomain chronyd[727]: Selected source 169.254.169.123
Jun 28 16:00:42 localhost.localdomain chronyd[727]: System clock TAI offset set to 37 seconds
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Source 167.179.96.146 offline
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Source 103.202.216.35 offline
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Source 129.250.35.250 offline
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Source 194.0.5.123 offline
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Source 169.254.169.123 offline
Jun 28 16:19:33 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[727]: Can't synchronise: no selectable sources
[root@ip-172-31-4-206 ~]# service chronyd restart
Redirecting to /bin/systemctl restart chronyd.service
[root@ip-172-31-4-206 ~]#
[root@ip-172-31-4-206 ~]#
[root@ip-172-31-4-206 ~]# service chronyd status
Redirecting to /bin/systemctl status chronyd.service
· chronyd.service - NTP client/server
Loaded: loaded (/usr/lib/systemd/system/chronyd.service; enabled; vendor preset: enabled)
Active: active (running) since Sun 2020-06-28 16:34:41 UTC; 2s ago
Docs: man:chronyd(8)
man:chrony.conf(5)
Process: 25252 ExecStartPost=/usr/libexec/chrony-helper update-daemon (code=exited, status=0/SUCCESS)
Process: 25247 ExecStart=/usr/sbin/chronyd $OPTIONS (code=exited, status=0/SUCCESS)
Main PID: 25250 (chronyd)
Tasks: 1 (limit: 49683)
Memory: 1.0M
CGroup: /system.slice/chronyd.service
└─25250 /usr/sbin/chronyd
Jun 28 16:34:41 ip-172-31-4-206.ap-northeast-1.compute.internal systemd[1]: Starting NTP client/server...
Jun 28 16:34:41 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[25250]: chronyd version 3.5 starting (+CMDMON +NTP +REFCLOCK +RTC +PRIVDROP +SCFILTER +SIGND>
Jun 28 16:34:41 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[25250]: Frequency 35.298 +/- 0.039 ppm read from /var/lib/chrony/drift
Jun 28 16:34:41 ip-172-31-4-206.ap-northeast-1.compute.internal chronyd[25250]: Using right/UTC timezone to obtain leap second data
Jun 28 16:34:41 ip-172-31-4-206.ap-northeast-1.compute.internal systemd[1]: Started NTP client/server.
[root@ip-172-31-4-206 ~]# chronyc -n sources
210 Number of sources = 5
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
^* 169.254.169.123 3 4 17 4 -2369ns[ -27us] +/- 451us
^- 94.154.96.7 2 6 17 5 +30ms[ +30ms] +/- 148ms
^- 185.51.192.34 2 6 17 3 -2951us[-2951us] +/- 150ms
^- 188.125.64.6 2 6 17 3 +9526us[+9526us] +/- 143ms
^- 216.218.254.202 1 6 17 5 +15ms[ +15ms] +/- 72ms
[root@ip-172-31-4-206 ~]#
[root@ip-172-31-4-206 ~]# contrail-status
Pod Service Original Name Original Version State Id Status
rsyslogd nightly-master running 5fc76e57c156 Up 16 minutes
vrouter agent contrail-vrouter-agent nightly-master running bce023d8e6e0 Up 5 minutes
vrouter nodemgr contrail-nodemgr nightly-master running 9439a304cbcf Up 5 minutes
vrouter provisioner contrail-provisioner nightly-master running 1531b1403e49 Up 5 minutes
WARNING: container with original name '' have Pod or Service empty. Pod: '' / Service: 'rsyslogd'. Please pass NODE_TYPE with pod name to container's env
vrouter kernel module is PRESENT
== Contrail vrouter ==
nodemgr: active
agent: active
[root@ip-172-31-4-206 ~]#原文链接：
https://github.com/tnaganawa/tungstenfabric-docs/blob/master/TungstenFabricKnowledgeBase.md
往期精选
Tungsten Fabric知识库丨vRouter内部运行探秘
Tungsten Fabric知识库丨更多组件内部探秘
Tungsten Fabric知识库丨 构建、安装与公有云部署
Tungsten Fabric知识库丨测试2000个vRouter节点部署
Tungsten Fabric入门宝典系列文章——
- 首次启动和运行指南
- TF组件的七种“武器”
- 编排器集成
- 关于安装的那些事（上）
- 关于安装的那些事（下）
- 主流监控系统工具的集成
- 开始第二天的工作
- 8个典型故障及排查Tips
- 关于集群更新的那些事
- 说说L3V P N及EV P N 集成
11.关于服务链、BGPaaS及其它
12.关于多集群和多数据中心
13.多编排器用法及配置
