查看官方文檔升級的操作需要做如下注意事項。
創(chuàng)新互聯(lián)建站是專業(yè)的天橋網(wǎng)站建設公司，天橋接單;提供成都網(wǎng)站制作、網(wǎng)站建設,網(wǎng)頁設計,網(wǎng)站設計,建網(wǎng)站,PHP網(wǎng)站建設等專業(yè)做網(wǎng)站服務;采用PHP框架,可快速的進行天橋網(wǎng)站開發(fā)網(wǎng)頁制作和功能擴展;專業(yè)做搜索引擎喜愛的網(wǎng)站,專業(yè)的做網(wǎng)站團隊,希望更多企業(yè)前來合作!
所以針對現(xiàn)有的情況，需要先升級至 2.6.5+ ，再升級 3.x。
2019/12/25
現(xiàn)有環(huán)境，使用 etcdv2 進行存儲的 calico 數(shù)據(jù)。
重慶分公司
重慶分公司
[root@k8s-1 kubelet]# which etcdv2
alias etcdv2='export ETCDCTL_API=2; /bin/etcdctl --ca-file /etc/etcd/ssl/etcd-root-ca.pem --cert-file /etc/etcd/ssl/etcd.pem --key-file /etc/etcd/ssl/etcd-key.pem --endpoints https://10.111.32.239:2379,https://10.111.32.241:2379,https://10.111.32.242:2379'
[root@k8s-1 kubelet]# etcdv2 ls /calico/ipam/v2/assignment/ipv4
/calico/ipam/v2/assignment/ipv4/block
[root@k8s-1 kubelet]# etcdv2 ls /calico/ipam/v2/assignment/ipv4/block
/calico/ipam/v2/assignment/ipv4/block/10.20.134.64-26
/calico/ipam/v2/assignment/ipv4/block/10.20.253.64-26
/calico/ipam/v2/assignment/ipv4/block/10.20.28.192-26
/calico/ipam/v2/assignment/ipv4/block/10.20.51.128-26
/calico/ipam/v2/assignment/ipv4/block/10.20.78.0-26
/calico/ipam/v2/assignment/ipv4/block/10.20.112.64-26
/calico/ipam/v2/assignment/ipv4/block/10.20.15.128-26
/calico/ipam/v2/assignment/ipv4/block/10.20.235.0-26
/calico/ipam/v2/assignment/ipv4/block/10.20.53.64-26
/calico/ipam/v2/assignment/ipv4/block/10.20.72.128-26
根據(jù)文檔中的說明，升級至 3.0 需要至少 2.6.5+ ,且需要進行一些手動的操作，因為 3.x 的使用 etcdv3, 而 2.6.x 的使用 etcdv2。
現(xiàn)在集群使用的是 2.6.1 的版本，先將其升級至 2.6.5+。
下載 calico.yaml 文件
[root@docker-182 v2.6]# wget https://docs.projectcalico.org/v2.6/getting-started/kubernetes/installation/rbac.yaml
[root@docker-182 v2.6]# wget https://docs.projectcalico.org/v2.6/getting-started/kubernetes/installation/hosted/calico.yaml
# 更改 calico.yaml 中的配置
[root@docker-182 v2.6]# sh -x modify_calico_yaml.sh[root@docker-182 v2.6]# grep image calico.yaml
image: quay.io/calico/node:v2.6.12
image: quay.io/calico/cni:v1.11.8
image: quay.io/calico/kube-controllers:v1.0.5
image: quay.io/calico/kube-controllers:v1.0.5
文檔中說的一些升級步驟，比如先升級 calico-kube-controllers ,再升級 calico-node 的daemonset ,這里就直接 apply 新的資源文件
并不包含 calico 的 rbac 資源。
[root@docker-182 v2.6]# k239 apply -f calico.yaml
configmap "calico-config" unchanged
secret "calico-etcd-secrets" unchanged
daemonset "calico-node" configured
deployment "calico-kube-controllers" configured
deployment "calico-policy-controller" configured
serviceaccount "calico-kube-controllers" unchanged
serviceaccount "calico-node" unchanged
提交之后， daemonset 的 calico-node 并沒有更新，現(xiàn)在刪除 pod ，使其更新
[root@k8s-1 v2.6]# kubectl -n kube-system get pod -o wide |grep calico
calico-kube-controllers-6768b96c5f-rdbjp 1/1 Running 0 4m 10.111.32.243 k8s-4.geotmt.com
calico-node-45lnh 0/1 ContainerCreating 0 4h 10.111.32.241 k8s-2.geotmt.com
calico-node-49mq7 1/1 Running 1 5h 10.111.32.243 k8s-4.geotmt.com
calico-node-m86hr 1/1 Running 0 5h 10.111.32.244 k8s-5.geotmt.com
calico-node-mm5fz 0/1 ContainerCreating 0 4h 10.111.32.239 k8s-1.geotmt.com
calico-node-shrfw 1/1 Running 0 4h 10.111.32.242 k8s-3.geotmt.com
calico-node-xx8hk 1/1 Running 0 5h 10.111.32.245 k8s-6.geotmt.com
其中一個的示例，新的 calico-node 其中有兩個容器。
[root@k8s-1 v2.6]# kubectl -n kube-system get pod -o wide |grep calico |grep k8s-6
calico-node-fj4t8 2/2 Running 0 25s 10.111.32.245 k8s-6.geotmt.com測試 ping 其他節(jié)點的 pod 正常
bash-4.4# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
2: tunl0@NONE: mtu 1480 qdisc noop state DOWN qlen 1000
link/ipip 0.0.0.0 brd 0.0.0.0
4: eth0@if30: mtu 1500 qdisc noqueue state UP
link/ether 6e:20:a3:45:42:49 brd ff:ff:ff:ff:ff:ff
inet 10.20.235.12/32 scope global eth0
valid_lft forever preferred_lft forever
bash-4.4# ping 10.20.15.135
PING 10.20.15.135 (10.20.15.135): 56 data bytes
64 bytes from 10.20.15.135: seq=0 ttl=62 time=1.133 ms
64 bytes from 10.20.15.135: seq=1 ttl=62 time=0.631 ms
這個版本的仍需手動添加 toleration，以便在 master 節(jié)點上部署 pod。
升級至 2.6.12 完成。
上述兩條都滿足。
[root@k8s-1 net.d]# etcdctl version
etcdctl version: 3.3.11
API version: 3.3
[root@docker-182 ansible]# wget https://github.com/projectcalico/calico-upgrade/releases/download/v1.0.5/calico-upgrade
[root@docker-182 k8s_239]# ansible-playbook install_calico-upgrade.yml
使用 dry-run 執(zhí)行測試
[root@k8s-1 calico-upgrade]# calico-upgrade dry-run --output-dir=tmp --apiconfigv1 /etc/calico/apiconfigv1.cfg --apiconfigv3 /etc/calico/apiconfigv3.cfg
[root@k8s-1 calico-upgrade]# calico-upgrade start --ignore-v3-data --apiconfigv1 /etc/calico/apiconfigv1.cfg --apiconfigv3 /etc/calico/apiconfigv3.cfg
Preparing reports directory
* creating report directory if it does not exist
* validating permissions and removing old reports
Checking Calico version is suitable for migration
* determined Calico version of: v2.6.12
* the v1 API data can be migrated to the v3 API
Validating conversion of v1 data to v3
* handling FelixConfiguration (global) resource
* handling ClusterInformation (global) resource
* handling FelixConfiguration (per-node) resources
* handling BGPConfiguration (global) resource
* handling Node resources
* handling BGPPeer (global) resources
* handling BGPPeer (node) resources
* handling HostEndpoint resources
* handling IPPool resources
* handling GlobalNetworkPolicy resources
* handling Profile resources
* handling WorkloadEndpoint resources
* data conversion successful
Data conversion validated successfully
Validating the v3 datastore
* the v3 datastore is not empty
-------------------------------------------------------------------------------
Successfully validated v1 to v3 conversion.
You are about to start the migration of Calico v1 data format to Calico v3 data
format. During this time and until the upgrade is completed Calico networking
will be paused - which means no new Calico networked endpoints can be created.
No Calico configuration should be modified using calicoctl during this time.
Type "yes" to proceed (any other input cancels): yes
Pausing Calico networking
* successfully paused Calico networking in the v1 configuration
Calico networking is now paused - waiting for 15s
Querying current v1 snapshot and converting to v3
* handling FelixConfiguration (global) resource
* handling ClusterInformation (global) resource
* handling FelixConfiguration (per-node) resources
* handling BGPConfiguration (global) resource
* handling Node resources
* handling BGPPeer (global) resources
* handling BGPPeer (node) resources
* handling HostEndpoint resources
* handling IPPool resources
* handling GlobalNetworkPolicy resources
* handling Profile resources
* handling WorkloadEndpoint resources
* data converted successfully
Storing v3 data
* Storing resources in v3 format
* success: resources stored in v3 datastore
Migrating IPAM data
* listing and converting IPAM allocation blocks
* listing and converting IPAM affinity blocks
* listing IPAM handles
* storing IPAM data in v3 format
* IPAM data migrated successfully
Data migration from v1 to v3 successful
* check the output for details of the migrated resources
* continue by upgrading your calico/node versions to Calico v3.x
-------------------------------------------------------------------------------
Successfully migrated Calico v1 data to v3 format.
Follow the detailed upgrade instructions available in the release documentation
to complete the upgrade. This includes:
* upgrading your calico/node instances and orchestrator plugins (e.g. CNI) to
the required v3.x release
* running 'calico-upgrade complete' to complete the upgrade and resume Calico
networking
See report(s) below for details of the migrated data.
Reports:
- name conversion: /root/calico-upgrade/calico-upgrade-report/convertednames
[root@docker-182 v3.0]# wget https://docs.projectcalico.org/v3.0/getting-started/kubernetes/installation/rbac.yaml
[root@docker-182 v3.0]# wget https://docs.projectcalico.org/v3.0/getting-started/kubernetes/installation/hosted/calico.yaml
3.0的改變可參考 3.0release note
預先下載所需鏡像
[root@docker-182 v3.0]# grep image calico.yaml
image: quay.io/calico/node:v3.0.12
image: quay.io/calico/cni:v3.0.12
image: quay.io/calico/kube-controllers:v3.0.12
[root@docker-182 v3.0]# k239 apply -f calico.yaml
configmap "calico-config" configured
secret "calico-etcd-secrets" unchanged
daemonset "calico-node" configured
deployment "calico-kube-controllers" configured
serviceaccount "calico-kube-controllers" unchanged
serviceaccount "calico-node" unchanged這里的 pod 可以實現(xiàn)滾動重啟，待pod 都升級完成后。
[root@k8s-1 calico-upgrade]# calico-upgrade complete --apiconfigv1 /etc/calico/apiconfigv1.cfg --apiconfigv3 /etc/calico/apiconfigv3.cfg
You are about to complete the upgrade process to Calico v3. At this point, the
v1 format data should have been successfully converted to v3 format, and all
calico/node instances and orchestrator plugins (e.g. CNI) should be running
Calico v3.x.
Type "yes" to proceed (any other input cancels): yes
Completing upgrade
Enabling Calico networking for v3
* successfully resumed Calico networking in the v3 configuration (updated
ClusterInformation)
Upgrade completed successfully
-------------------------------------------------------------------------------
Successfully completed the upgrade process.
如不執(zhí)行上述命令,會有如下報錯
E1225 19:56:04.837028 3281 kuberuntime_sandbox.go:54] CreatePodSandbox for pod "demo-deployment-6f4c6779b-b8zqq_default(1dd28cf0-270d-11ea-bd6c-c6a864ab864a)" failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to set up pod "demo-deployment-6f4c6779b-b8zqq_default" network: Calico is currently not ready to process requests
E1225 19:56:04.837049 3281 kuberuntime_manager.go:647] createPodSandbox for pod "demo-deployment-6f4c6779b-b8zqq_default(1dd28cf0-270d-11ea-bd6c-c6a864ab864a)" failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to set up pod "demo-deployment-6f4c6779b-b8zqq_default" network: Calico is currently not ready to process requests
E1225 19:56:04.837167 3281 pod_workers.go:186] Error syncing pod 1dd28cf0-270d-11ea-bd6c-c6a864ab864a ("demo-deployment-6f4c6779b-b8zqq_default(1dd28cf0-270d-11ea-bd6c-c6a864ab864a)"), skipping: failed to "CreatePodSandbox" for "demo-deployment-6f4c6779b-b8zqq_default(1dd28cf0-270d-11ea-bd6c-c6a864ab864a)" with CreatePodSandboxError: "CreatePodSandbox for pod \"demo-deployment-6f4c6779b-b8zqq_default(1dd28cf0-270d-11ea-bd6c-c6a864ab864a)\" failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to set up pod \"demo-deployment-6f4c6779b-b8zqq_default\" network: Calico is currently not ready to process requests"
升級至 3.0.12 成功。
根據(jù) 3.11 的 Upgrading Calico on Kubernetes 說明。升級時，只需要提交新的資源文件即可(本環(huán)境不涉及 Application Layer Policy)。
這個版本的 calico 已經(jīng)可以完整支持 k8s api 的datastore， 更新時要注意下載文件時是否與自己的環(huán)境契合。
本環(huán)境下載 etcd datastore 的版本。
[root@docker-182 v3.11]# wget https://docs.projectcalico.org/v3.11/manifests/calico-etcd.yaml
# 修改其中關于 etcd 的配置
[root@docker-182 v3.11]# bash -x modify_calico_yaml.sh
[root@docker-182 v3.11]# grep image calico-etcd.yaml
image: calico/cni:v3.11.1
image: calico/pod2daemon-flexvol:v3.11.1
image: calico/node:v3.11.1
image: calico/kube-controllers:v3.11.1[root@docker-182 v3.11]# k239 apply -f calico-etcd.yaml
secret "calico-etcd-secrets" unchanged
configmap "calico-config" configured
clusterrole "calico-kube-controllers" configured
clusterrolebinding "calico-kube-controllers" configured
clusterrole "calico-node" configured
clusterrolebinding "calico-node" configured
daemonset "calico-node" configured
serviceaccount "calico-node" unchanged
deployment "calico-kube-controllers" configured
serviceaccount "calico-kube-controllers" unchanged查看新版本的 pod, 每個 pod 內只有一個容器，這個版本的將 install-cni 和 flexvol-driver(舊版本沒有) 作為了 initContainers ，所以常駐的就只有一個容器了
[root@docker-182 ~]# k239 -n kube-system get pod -o wide |grep calico
calico-kube-controllers-85dc4fd46b-4wnmt 1/1 Running 0 1m 10.111.32.243 k8s-4.geotmt.com
calico-node-4bgkc 1/1 Running 0 59s 10.111.32.241 k8s-2.geotmt.com
calico-node-5jg2t 1/1 Running 0 31s 10.111.32.244 k8s-5.geotmt.com
calico-node-9fn6r 1/1 Running 0 43s 10.111.32.245 k8s-6.geotmt.com
calico-node-9n7dn 1/1 Running 0 1m 10.111.32.243 k8s-4.geotmt.com
calico-node-fxr46 1/1 Running 0 1m 10.111.32.239 k8s-1.geotmt.com
calico-node-pgh6c 1/1 Running 0 1m 10.111.32.242 k8s-3.geotmt.com測試 pod 的跨主機通信
[root@k8s-1 ~]# kubectl exec -it demo-deployment-6f4c6779b-b8zqq /bin/bash
bash-4.4# ping 10.20.235.12
PING 10.20.235.12 (10.20.235.12): 56 data bytes
64 bytes from 10.20.235.12: seq=0 ttl=62 time=1.232 ms
^C
--- 10.20.235.12 ping statistics ---
1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max = 1.232/1.232/1.232 ms
bash-4.4# ping 10.20.253.80
PING 10.20.253.80 (10.20.253.80): 56 data bytes
64 bytes from 10.20.253.80: seq=0 ttl=62 time=1.730 ms
64 bytes from 10.20.253.80: seq=1 ttl=62 time=1.385 ms
^C
--- 10.20.253.80 ping statistics ---
2 packets transmitted, 2 packets received, 0% packet loss
round-trip min/avg/max = 1.385/1.557/1.730 ms
bash-4.4# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
2: tunl0@NONE: mtu 1480 qdisc noop state DOWN qlen 1000
link/ipip 0.0.0.0 brd 0.0.0.0
4: eth0@if51: mtu 1500 qdisc noqueue state UP
link/ether fa:d1:55:42:ab:6c brd ff:ff:ff:ff:ff:ff
inet 10.20.15.163/32 scope global eth0
valid_lft forever preferred_lft forever
測試pod重建分配地址,成功
[root@k8s-1 ~]# kubectl delete pod nginx-deployment-7b66d98974-2rh87
pod "nginx-deployment-7b66d98974-2rh87" deleted
[root@k8s-1 ~]# kubectl get pod nginx-deployment-7b66d98974-nd8h7 -o wide
NAME READY STATUS RESTARTS AGE IP NODE
nginx-deployment-7b66d98974-nd8h7 1/1 Running 0 1m 10.20.253.86 k8s-4.geotmt.comcalico 3.0.12 升級至 3.11.1 成功。
