K8s 新手。到目前为止，我有以下内容:
- docker-ce-19.03.8
- docker-ce-cli-19.03.8
- containerd.io-1.2.13
- kubelet-1.18.5
- kubeadm-1.18.5
- kubectl-1.18.5
- etcd-3.4.10
- 使用 Flannel 作为 Pod Overlay Net
- 执行了所有主机级工作(SELinux 许可、交换等)
- 本地 Vsphere 环境中的所有 Centos7 (6.7U3)
我已经构建了所有配置，目前拥有:
- 具有点对点和客户端-服务器加密传输功能的 3 节点外部/独立 etcd 集群
- 一个 3 节点控制平面集群 - kubeadm init 使用 x509 进行引导，并以 3 个 etcd 为目标(因此堆叠的 etcd 永远不会发生)
- HAProxy 和 Keepalived 安装在两个 etcd 集群成员上，对控制平面 (TCP6443) 上的 API 服务器端点进行负载平衡访问
- 6 个工作节点
- 使用树内 Vmware Cloud Provider 配置的存储(我知道它已弃用)——是的，这是我的默认 SC
状态检查:
- kubectl 集群信息报告:
[me@km-01 pods]$ kubectl cluster-info Kubernetes master is running at https://k8snlb:6443 KubeDNS is running at https://k8snlb:6443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
kubectl 获取所有 --all-namespaces 报告:
[me@km-01 pods]$ kubectl get all --all-namespaces -owide
NAMESPACE NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
ag1 pod/mssql-operator-68bcc684c4-rbzvn 1/1 Running 0 86m 10.10.4.133 kw-02.bogus.local <none> <none>
kube-system pod/coredns-66bff467f8-k6m94 1/1 Running 4 20h 10.10.0.11 km-01.bogus.local <none> <none>
kube-system pod/coredns-66bff467f8-v848r 1/1 Running 4 20h 10.10.0.10 km-01.bogus.local <none> <none>
kube-system pod/kube-apiserver-km-01.bogus.local 1/1 Running 8 10h x.x.x..25 km-01.bogus.local <none> <none>
kube-system pod/kube-controller-manager-km-01.bogus.local 1/1 Running 2 10h x.x.x..25 km-01.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-7l76c 1/1 Running 0 10h x.x.x..30 kw-01.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-8kft7 1/1 Running 0 10h x.x.x..33 kw-04.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-r5kqv 1/1 Running 0 10h x.x.x..34 kw-05.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-t6xcd 1/1 Running 0 10h x.x.x..35 kw-06.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-vhnx8 1/1 Running 0 10h x.x.x..32 kw-03.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-xdk2n 1/1 Running 0 10h x.x.x..31 kw-02.bogus.local <none> <none>
kube-system pod/kube-flannel-ds-amd64-z4kfk 1/1 Running 4 20h x.x.x..25 km-01.bogus.local <none> <none>
kube-system pod/kube-proxy-49hsl 1/1 Running 0 10h x.x.x..35 kw-06.bogus.local <none> <none>
kube-system pod/kube-proxy-62klh 1/1 Running 0 10h x.x.x..34 kw-05.bogus.local <none> <none>
kube-system pod/kube-proxy-64d5t 1/1 Running 0 10h x.x.x..30 kw-01.bogus.local <none> <none>
kube-system pod/kube-proxy-6ch42 1/1 Running 4 20h x.x.x..25 km-01.bogus.local <none> <none>
kube-system pod/kube-proxy-9css4 1/1 Running 0 10h x.x.x..32 kw-03.bogus.local <none> <none>
kube-system pod/kube-proxy-hgrx8 1/1 Running 0 10h x.x.x..33 kw-04.bogus.local <none> <none>
kube-system pod/kube-proxy-ljlsh 1/1 Running 0 10h x.x.x..31 kw-02.bogus.local <none> <none>
kube-system pod/kube-scheduler-km-01.bogus.local 1/1 Running 5 20h x.x.x..25 km-01.bogus.local <none> <none>
NAMESPACE NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE SELECTOR
ag1 service/ag1-primary NodePort 10.104.183.81 x.x.x..30,x.x.x..31,x.x.x..32,x.x.x..33,x.x.x..34,x.x.x..35 1433:30405/TCP 85m role.ag.mssql.microsoft.com/ag1=primary,type=sqlservr
ag1 service/ag1-secondary NodePort 10.102.52.31 x.x.x..30,x.x.x..31,x.x.x..32,x.x.x..33,x.x.x..34,x.x.x..35 1433:30713/TCP 85m role.ag.mssql.microsoft.com/ag1=secondary,type=sqlservr
ag1 service/mssql1 NodePort 10.96.166.108 x.x.x..30,x.x.x..31,x.x.x..32,x.x.x..33,x.x.x..34,x.x.x..35 1433:32439/TCP 86m name=mssql1,type=sqlservr
ag1 service/mssql2 NodePort 10.109.146.58 x.x.x..30,x.x.x..31,x.x.x..32,x.x.x..33,x.x.x..34,x.x.x..35 1433:30636/TCP 86m name=mssql2,type=sqlservr
ag1 service/mssql3 NodePort 10.101.234.186 x.x.x..30,x.x.x..31,x.x.x..32,x.x.x..33,x.x.x..34,x.x.x..35 1433:30862/TCP 86m name=mssql3,type=sqlservr
default service/kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 23h <none>
kube-system service/kube-dns ClusterIP 10.96.0.10 <none> 53/UDP,53/TCP,9153/TCP 20h k8s-app=kube-dns
NAMESPACE NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE CONTAINERS IMAGES SELECTOR
kube-system daemonset.apps/kube-flannel-ds-amd64 7 7 7 7 7 <none> 20h kube-flannel quay.io/coreos/flannel:v0.12.0-amd64 app=flannel
kube-system daemonset.apps/kube-flannel-ds-arm 0 0 0 0 0 <none> 20h kube-flannel quay.io/coreos/flannel:v0.12.0-arm app=flannel
kube-system daemonset.apps/kube-flannel-ds-arm64 0 0 0 0 0 <none> 20h kube-flannel quay.io/coreos/flannel:v0.12.0-arm64 app=flannel
kube-system daemonset.apps/kube-flannel-ds-ppc64le 0 0 0 0 0 <none> 20h kube-flannel quay.io/coreos/flannel:v0.12.0-ppc64le app=flannel
kube-system daemonset.apps/kube-flannel-ds-s390x 0 0 0 0 0 <none> 20h kube-flannel quay.io/coreos/flannel:v0.12.0-s390x app=flannel
kube-system daemonset.apps/kube-proxy 7 7 7 7 7 kubernetes.io/os=linux 20h kube-proxy k8s.gcr.io/kube-proxy:v1.18.7 k8s-app=kube-proxy
NAMESPACE NAME READY UP-TO-DATE AVAILABLE AGE CONTAINERS IMAGES SELECTOR
ag1 deployment.apps/mssql-operator 1/1 1 1 86m mssql-operator mcr.microsoft.com/mssql/ha:2019-CTP2.1-ubuntu app=mssql-operator
kube-system deployment.apps/coredns 2/2 2 2 20h coredns k8s.gcr.io/coredns:1.6.7 k8s-app=kube-dns
NAMESPACE NAME DESIRED CURRENT READY AGE CONTAINERS IMAGES SELECTOR
ag1 replicaset.apps/mssql-operator-68bcc684c4 1 1 1 86m mssql-operator mcr.microsoft.com/mssql/ha:2019-CTP2.1-ubuntu app=mssql-operator,pod-template-hash=68bcc684c4
kube-system replicaset.apps/coredns-66bff467f8 2 2 2 20h coredns k8s.gcr.io/coredns:1.6.7 k8s-app=kube-dns,pod-template-hash=66bff467f8
问题:有很多文章讨论 SQL2019 HA 构建。然而，似乎每一个都在云端，而我的是在 Vsphere 环境中的本地部署。它们看起来非常简单:按以下顺序运行 3 个脚本:operator.yaml、sql.yaml 和 ag-service.yaml。
对于事后实际截图环境的博客，应该至少有 7 个 pod(1 个 Operator、3 个 SQL Init、3 个 SQL)。如果你看看我前面提到的 all --all-namespaces 输出，我拥有一切(并且处于运行状态)，但除了正在运行的 Operator 之外没有 pod...???
我实际上将控制平面恢复为单节点只是为了尝试隔离日志。/var/log/container/* 和/var/log/pod/* 不包含任何有值(value)的内容来指示存储问题或 Pod 不存在的任何其他原因。可能还值得注意的是，我开始使用最新的 sql2019 标签:2019-latest，但是当我在那里得到相同的行为时，我决定尝试使用旧的位，因为很多博客都基于 CTP 2.1。
我可以使用 VCP 存储提供程序创建 PV 和 PVC。我有我的 Secrets，并且可以在 Secrets 商店中看到它们。
我无法解释为什么 Pod 丢失，或者在检查了 Journalctl、守护进程本身和/var/log 后去哪里查看，而且我没有看到任何迹象表明有人试图创建它们 -我改编的 kubectl apply -f mssql-server2019.yaml 运行完成，并且没有错误指示创建了 3 个 sql 对象和 3 个 sql 服务。但无论如何，这是针对 CTP2.1 的文件:
cat << EOF > mssql-server2019.yaml
apiVersion: mssql.microsoft.com/v1
kind: SqlServer
metadata:
labels: {name: mssql1, type: sqlservr}
name: mssql1
namespace: ag1
spec:
acceptEula: true
agentsContainerImage: mcr.microsoft.com/mssql/ha:2019-CTP2.1
availabilityGroups: [ag1]
instanceRootVolumeClaimTemplate:
accessModes: [ReadWriteOnce]
resources:
requests: {storage: 5Gi}
storageClass: default
saPassword:
secretKeyRef: {key: sapassword, name: sql-secrets}
sqlServerContainer: {image: 'mcr.microsoft.com/mssql/server:2019-CTP2.1'}
---
apiVersion: v1
kind: Service
metadata: {name: mssql1, namespace: ag1}
spec:
ports:
- {name: tds, port: 1433}
selector: {name: mssql1, type: sqlservr}
type: NodePort
externalIPs:
- x.x.x.30
- x.x.x.31
- x.x.x.32
- x.x.x.33
- x.x.x.34
- x.x.x.35
---
apiVersion: mssql.microsoft.com/v1
kind: SqlServer
metadata:
labels: {name: mssql2, type: sqlservr}
name: mssql2
namespace: ag1
spec:
acceptEula: true
agentsContainerImage: mcr.microsoft.com/mssql/ha:2019-CTP2.1
availabilityGroups: [ag1]
instanceRootVolumeClaimTemplate:
accessModes: [ReadWriteOnce]
resources:
requests: {storage: 5Gi}
storageClass: default
saPassword:
secretKeyRef: {key: sapassword, name: sql-secrets}
sqlServerContainer: {image: 'mcr.microsoft.com/mssql/server:2019-CTP2.1'}
---
apiVersion: v1
kind: Service
metadata: {name: mssql2, namespace: ag1}
spec:
ports:
- {name: tds, port: 1433}
selector: {name: mssql2, type: sqlservr}
type: NodePort
externalIPs:
- x.x.x.30
- x.x.x.31
- x.x.x.32
- x.x.x.33
- x.x.x.34
- x.x.x.35
---
apiVersion: mssql.microsoft.com/v1
kind: SqlServer
metadata:
labels: {name: mssql3, type: sqlservr}
name: mssql3
namespace: ag1
spec:
acceptEula: true
agentsContainerImage: mcr.microsoft.com/mssql/ha:2019-CTP2.1
availabilityGroups: [ag1]
instanceRootVolumeClaimTemplate:
accessModes: [ReadWriteOnce]
resources:
requests: {storage: 5Gi}
storageClass: default
saPassword:
secretKeyRef: {key: sapassword, name: sql-secrets}
sqlServerContainer: {image: 'mcr.microsoft.com/mssql/server:2019-CTP2.1'}
---
apiVersion: v1
kind: Service
metadata: {name: mssql3, namespace: ag1}
spec:
ports:
- {name: tds, port: 1433}
selector: {name: mssql3, type: sqlservr}
type: NodePort
externalIPs:
- x.x.x.30
- x.x.x.31
- x.x.x.32
- x.x.x.33
- x.x.x.34
- x.x.x.35
---
EOF
编辑1:kubectl日志-n ag mssql-operator-*
[sqlservers] 2020/08/14 14:36:48 Creating custom resource definition
[sqlservers] 2020/08/14 14:36:48 Created custom resource definition
[sqlservers] 2020/08/14 14:36:48 Waiting for custom resource definition to be available
[sqlservers] 2020/08/14 14:36:49 Watching for resources...
[sqlservers] 2020/08/14 14:37:08 Creating ConfigMap sql-operator
[sqlservers] 2020/08/14 14:37:08 Updating mssql1 in namespace ag1 ...
[sqlservers] 2020/08/14 14:37:08 Creating ConfigMap ag1
[sqlservers] ERROR: 2020/08/14 14:37:08 could not process update request: error creating ConfigMap ag1: v1.ConfigMap: ObjectMeta: v1.ObjectMeta: readObjectFieldAsBytes: expect : after object field, parsing 627 ...:{},"k:{\"... at {"kind":"ConfigMap","apiVersion":"v1","metadata":{"name":"ag1","namespace":"ag1","selfLink":"/api/v1/namespaces/ag1/configmaps/ag1","uid":"33af6232-4464-4290-bb14-b21e8f72e361","resourceVersion":"314186","creationTimestamp":"2020-08-14T14:37:08Z","ownerReferences":[{"apiVersion":"mssql.microsoft.com/v1","kind":"ReplicationController","name":"mssql1","uid":"e71a7246-2776-4d96-9735-844ee136a37d","controller":false}],"managedFields":[{"manager":"mssql-server-k8s-operator","operation":"Update","apiVersion":"v1","time":"2020-08-14T14:37:08Z","fieldsType":"FieldsV1","fieldsV1":{"f:metadata":{"f:ownerReferences":{".":{},"k:{\"uid\":\"e71a7246-2776-4d96-9735-844ee136a37d\"}":{".":{},"f:apiVersion":{},"f:controller":{},"f:kind":{},"f:name":{},"f:uid":{}}}}}}]}}
[sqlservers] 2020/08/14 14:37:08 Updating ConfigMap sql-operator
[sqlservers] 2020/08/14 14:37:08 Updating mssql2 in namespace ag1 ...
[sqlservers] ERROR: 2020/08/14 14:37:08 could not process update request: error getting ConfigMap ag1: v1.ConfigMap: ObjectMeta: v1.ObjectMeta: readObjectFieldAsBytes: expect : after object field, parsing 627 ...:{},"k:{\"... at {"kind":"ConfigMap","apiVersion":"v1","metadata":{"name":"ag1","namespace":"ag1","selfLink":"/api/v1/namespaces/ag1/configmaps/ag1","uid":"33af6232-4464-4290-bb14-b21e8f72e361","resourceVersion":"314186","creationTimestamp":"2020-08-14T14:37:08Z","ownerReferences":[{"apiVersion":"mssql.microsoft.com/v1","kind":"ReplicationController","name":"mssql1","uid":"e71a7246-2776-4d96-9735-844ee136a37d","controller":false}],"managedFields":[{"manager":"mssql-server-k8s-operator","operation":"Update","apiVersion":"v1","time":"2020-08-14T14:37:08Z","fieldsType":"FieldsV1","fieldsV1":{"f:metadata":{"f:ownerReferences":{".":{},"k:{\"uid\":\"e71a7246-2776-4d96-9735-844ee136a37d\"}":{".":{},"f:apiVersion":{},"f:controller":{},"f:kind":{},"f:name":{},"f:uid":{}}}}}}]}}
[sqlservers] 2020/08/14 14:37:08 Updating ConfigMap sql-operator
[sqlservers] 2020/08/14 14:37:08 Updating mssql3 in namespace ag1 ...
[sqlservers] ERROR: 2020/08/14 14:37:08 could not process update request: error getting ConfigMap ag1: v1.ConfigMap: ObjectMeta: v1.ObjectMeta: readObjectFieldAsBytes: expect : after object field, parsing 627 ...:{},"k:{\"... at {"kind":"ConfigMap","apiVersion":"v1","metadata":{"name":"ag1","namespace":"ag1","selfLink":"/api/v1/namespaces/ag1/configmaps/ag1","uid":"33af6232-4464-4290-bb14-b21e8f72e361","resourceVersion":"314186","creationTimestamp":"2020-08-14T14:37:08Z","ownerReferences":[{"apiVersion":"mssql.microsoft.com/v1","kind":"ReplicationController","name":"mssql1","uid":"e71a7246-2776-4d96-9735-844ee136a37d","controller":false}],"managedFields":[{"manager":"mssql-server-k8s-operator","operation":"Update","apiVersion":"v1","time":"2020-08-14T14:37:08Z","fieldsType":"FieldsV1","fieldsV1":{"f:metadata":{"f:ownerReferences":{".":{},"k:{\"uid\":\"e71a7246-2776-4d96-9735-844ee136a37d\"}":{".":{},"f:apiVersion":{},"f:controller":{},"f:kind":{},"f:name":{},"f:uid":{}}}}}}]}}
我检查了我的运算符和 mssql2019.yamls(特别是围绕 SqlServer 类型，因为这似乎是它失败的地方)，并且无法识别任何明显的不一致或差异。
最佳答案
所以你的操作符正在运行:
ag1 pod/pod/mssql-operator-68bcc684c4-rbzvn 1/1 Running 0 86m 10.10.4.133 kw-02.bogus.local <none> <none>
我首先查看那里的日志:
kubectl -n ag1 logs pod/mssql-operator-68bcc684c4-rbzvn
很可能它需要与云提供商(即 Azure)交互，并且不支持 VMware，但请检查日志内容👀。
更新:
根据您发布的日志，您似乎使用的是 K8s 1.18 并且运算符不兼容。它正在尝试使用 kube-apiserver 拒绝的规范创建 ConfigMap。
✌️</p>
关于docker - K8s SQL2019 HA 容器 - 老兄，我的 Pod 在哪里？，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/63405840/