原文：https://medium.com/better-programming/a-closer-look-at-etcd-the-brain-of-a-kubernetes-cluster-788c8ea759a5

Etcd是Kubernetes用于存储集群各种状态信息（配置信息，运行）一个很重要的组件，这篇文章，我们带领大家掀开Etcd的神秘面纱，理解他是如何存储这些各种各样的碎片信息的。

Etcd的概况

Etcd 是一个分布式的，依赖key-value存储的，最重要的分布式数据存储系统
--https://etcd.io

在Kubernetes的世界里面，etcd是服务发现，集群状态存储以及其配置的基石。

Etcd以集群部署，节点间通信是通过Raft算法处理。在生产环境中，集群包含至少3个节点。在 http://thesecretlivesofdata.com/ 中使用动画很好地地解释了Raft算法如何运作以及整个集群的生命周期，包含以下几点

leader节点的选举
日志的复制与同步

Kubernetes中的Etcd

在Kubernetes集群的背景下，etcd实例可以作为Pod被部署在master节点上。下面是我们这篇文章会使用的Kubernetes模型。

如果想增加安全校验与弹性伸缩的功能的话，也可以将etcd部署成一个内部集群。

下面这个时序图来自Hepio的博客，展示了当一个Pod被创建的时候，Kubernetes内部模块之间的处理流程，形象地阐述了API Server和etcd之间的相互作用。

Kubernetes测试集群

这个部分我们使用在DigitalOcean的kubeadm创建一个3节点的Kubernetes测试集群，选择weavenet作为网络附加组件。这个集群的etcd运行在master节点上。我们并没有配置一个真实环境的HA集群，但已经足够我们去探索etcd的数据存储功能了。

$ kubectl get nodes
NAME    STATUS ROLES  AGE   VERSION
node-01 Ready  master 56m   v1.15.2
node-02 Ready  <none> 2m17  v1.15.2
node-03 Ready  <none> 2m17  v1.15.2

Etcd Pod

首先，我们将集群中所有正在运行的Pods先列出来：

$ kubectl get pods --all-namespaces
NAMESPACE   NAME                           READY STATUS  RESTART AGE
kube-system coredns-5c98db65d4–5kjjv       1/1   Running 0       57m
kube-system coredns-5c98db65d4–88hkq       1/1   Running 0       57m
kube-system etcd-node-01                   1/1   Running 0       56m
kube-system kube-apiserver-node-01         1/1   Running 0       56m
kube-system kube-controller-manager-node-01 1/1  Running 0       56m
kube-system kube-proxy-7642v               1/1   Running 0       3m
kube-system kube-proxy-jsp4r               1/1   Running 0       3m
kube-system kube-proxy-xj8qm               1/1   Running 0       57m
kube-system kube-scheduler-node-01         1/1   Running 0       56m
kube-system weave-net-2hvbx                2/2   Running 0       87s
kube-system weave-net-5mrjl                2/2   Running 0       87s
kube-system weave-net-c76fx                2/2   Running 0       87s

当一个集群刚被初始化的时候，只有在namespace为kube-system中的Pod是运行状态的。这些Pod是用来作为集群的任务管理的。我们比较关心的Pod是etcd-node-01，他是一个ectd的实例，用于存储集群状态。

我们运行一个shell命令，进入到这个Pod里面并且查看这个ctcd container的配置

通过使用--advertise-client-urls标识我们可以拿到所有的key/value对，通过etcdctl命令将他们保存到etcd-kv.json中

$ ADVERTISE_URL="https://134.209.178.162:2379"

$ kubectl exec etcd-node-01 -n kube-system -- sh -c \
"ETCDCTL_API=3 etcdctl \
--endpoints $ADVERTISE_URL \
--cacert /etc/kubernetes/pki/etcd/ca.crt \
--key /etc/kubernetes/pki/etcd/server.key \
--cert /etc/kubernetes/pki/etcd/server.crt \
get \"\" --prefix=true -w json" > etcd-kv.json

我们快速查看一下这个文件，所有的key对应的values，并且所有都编码成了base64。

我们先获取所有的key到一个text文件里面看看他们都长什么样子，我将所有的key都列出来了，所以有点点长

$ for k in $(cat etcd-kv.json | jq '.kvs[].key' | cut -d '"' -f2); do echo $k | base64 --decode; echo; done

/registry/apiregistration.k8s.io/apiservices/v1.
/registry/apiregistration.k8s.io/apiservices/v1.apps
/registry/apiregistration.k8s.io/apiservices/v1.authentication.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.authorization.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.autoscaling
/registry/apiregistration.k8s.io/apiservices/v1.batch
/registry/apiregistration.k8s.io/apiservices/v1.coordination.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.networking.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.rbac.authorization.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.scheduling.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1.storage.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.admissionregistration.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.apiextensions.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.apps
/registry/apiregistration.k8s.io/apiservices/v1beta1.authentication.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.authorization.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.batch
/registry/apiregistration.k8s.io/apiservices/v1beta1.certificates.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.coordination.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.events.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.extensions
/registry/apiregistration.k8s.io/apiservices/v1beta1.networking.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.node.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.policy
/registry/apiregistration.k8s.io/apiservices/v1beta1.rbac.authorization.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.scheduling.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta1.storage.k8s.io
/registry/apiregistration.k8s.io/apiservices/v1beta2.apps
/registry/apiregistration.k8s.io/apiservices/v2beta1.autoscaling
/registry/apiregistration.k8s.io/apiservices/v2beta2.autoscaling
/registry/certificatesigningrequests/csr-h9mcg
/registry/certificatesigningrequests/csr-qwnxf
/registry/certificatesigningrequests/csr-xklls
/registry/clusterrolebindings/cluster-admin
/registry/clusterrolebindings/kubeadm:kubelet-bootstrap
/registry/clusterrolebindings/kubeadm:node-autoapprove-bootstrap
/registry/clusterrolebindings/kubeadm:node-autoapprove-certificate-rotation
/registry/clusterrolebindings/kubeadm:node-proxier
/registry/clusterrolebindings/system:basic-user
/registry/clusterrolebindings/system:controller:attachdetach-controller
/registry/clusterrolebindings/system:controller:certificate-controller
/registry/clusterrolebindings/system:controller:clusterrole-aggregation-controller
/registry/clusterrolebindings/system:controller:cronjob-controller
/registry/clusterrolebindings/system:controller:daemon-set-controller
/registry/clusterrolebindings/system:controller:deployment-controller
/registry/clusterrolebindings/system:controller:disruption-controller
/registry/clusterrolebindings/system:controller:endpoint-controller
/registry/clusterrolebindings/system:controller:expand-controller
/registry/clusterrolebindings/system:controller:generic-garbage-collector
/registry/clusterrolebindings/system:controller:horizontal-pod-autoscaler
/registry/clusterrolebindings/system:controller:job-controller
/registry/clusterrolebindings/system:controller:namespace-controller
/registry/clusterrolebindings/system:controller:node-controller
/registry/clusterrolebindings/system:controller:persistent-volume-binder
/registry/clusterrolebindings/system:controller:pod-garbage-collector
/registry/clusterrolebindings/system:controller:pv-protection-controller
/registry/clusterrolebindings/system:controller:pvc-protection-controller
/registry/clusterrolebindings/system:controller:replicaset-controller
/registry/clusterrolebindings/system:controller:replication-controller
/registry/clusterrolebindings/system:controller:resourcequota-controller
/registry/clusterrolebindings/system:controller:route-controller
/registry/clusterrolebindings/system:controller:service-account-controller
/registry/clusterrolebindings/system:controller:service-controller
/registry/clusterrolebindings/system:controller:statefulset-controller
/registry/clusterrolebindings/system:controller:ttl-controller
/registry/clusterrolebindings/system:coredns
/registry/clusterrolebindings/system:discovery
/registry/clusterrolebindings/system:kube-controller-manager
/registry/clusterrolebindings/system:kube-dns
/registry/clusterrolebindings/system:kube-scheduler
/registry/clusterrolebindings/system:node
/registry/clusterrolebindings/system:node-proxier
/registry/clusterrolebindings/system:public-info-viewer
/registry/clusterrolebindings/system:volume-scheduler
/registry/clusterrolebindings/weave-net
/registry/clusterroles/admin
/registry/clusterroles/cluster-admin
/registry/clusterroles/edit
/registry/clusterroles/system:aggregate-to-admin
/registry/clusterroles/system:aggregate-to-edit
/registry/clusterroles/system:aggregate-to-view
/registry/clusterroles/system:auth-delegator
/registry/clusterroles/system:basic-user
/registry/clusterroles/system:certificates.k8s.io:certificatesigningrequests:nodeclient
/registry/clusterroles/system:certificates.k8s.io:certificatesigningrequests:selfnodeclient
/registry/clusterroles/system:controller:attachdetach-controller
/registry/clusterroles/system:controller:certificate-controller
/registry/clusterroles/system:controller:clusterrole-aggregation-controller
/registry/clusterroles/system:controller:cronjob-controller
/registry/clusterroles/system:controller:daemon-set-controller
/registry/clusterroles/system:controller:deployment-controller
/registry/clusterroles/system:controller:disruption-controller
/registry/clusterroles/system:controller:endpoint-controller
/registry/clusterroles/system:controller:expand-controller
/registry/clusterroles/system:controller:generic-garbage-collector
/registry/clusterroles/system:controller:horizontal-pod-autoscaler
/registry/clusterroles/system:controller:job-controller
/registry/clusterroles/system:controller:namespace-controller
/registry/clusterroles/system:controller:node-controller
/registry/clusterroles/system:controller:persistent-volume-binder
/registry/clusterroles/system:controller:pod-garbage-collector
/registry/clusterroles/system:controller:pv-protection-controller
/registry/clusterroles/system:controller:pvc-protection-controller
/registry/clusterroles/system:controller:replicaset-controller
/registry/clusterroles/system:controller:replication-controller
/registry/clusterroles/system:controller:resourcequota-controller
/registry/clusterroles/system:controller:route-controller
/registry/clusterroles/system:controller:service-account-controller
/registry/clusterroles/system:controller:service-controller
/registry/clusterroles/system:controller:statefulset-controller
/registry/clusterroles/system:controller:ttl-controller
/registry/clusterroles/system:coredns
/registry/clusterroles/system:csi-external-attacher
/registry/clusterroles/system:csi-external-provisioner
/registry/clusterroles/system:discovery
/registry/clusterroles/system:heapster
/registry/clusterroles/system:kube-aggregator
/registry/clusterroles/system:kube-controller-manager
/registry/clusterroles/system:kube-dns
/registry/clusterroles/system:kube-scheduler
/registry/clusterroles/system:kubelet-api-admin
/registry/clusterroles/system:node
/registry/clusterroles/system:node-bootstrapper
/registry/clusterroles/system:node-problem-detector
/registry/clusterroles/system:node-proxier
/registry/clusterroles/system:persistent-volume-provisioner
/registry/clusterroles/system:public-info-viewer
/registry/clusterroles/system:volume-scheduler
/registry/clusterroles/view
/registry/clusterroles/weave-net
/registry/configmaps/kube-public/cluster-info
/registry/configmaps/kube-system/coredns
/registry/configmaps/kube-system/extension-apiserver-authentication
/registry/configmaps/kube-system/kube-proxy
/registry/configmaps/kube-system/kubeadm-config
/registry/configmaps/kube-system/kubelet-config-1.15
/registry/configmaps/kube-system/weave-net
/registry/controllerrevisions/kube-system/kube-proxy-84c6b844cd
/registry/controllerrevisions/kube-system/weave-net-7db89b6d4
/registry/daemonsets/kube-system/kube-proxy
/registry/daemonsets/kube-system/weave-net
/registry/deployments/kube-system/coredns
/registry/events/default/node-01.15b9e0cd75ea6932
/registry/events/default/node-02.15b9e0ae0342c323
/registry/events/default/node-02.15b9e0ae0f9c2ae3
/registry/events/default/node-02.15b9e0ae0f9c5fa9
/registry/events/default/node-02.15b9e0ae0f9c7206
/registry/events/default/node-02.15b9e0ae1575182e
/registry/events/default/node-02.15b9e0aea1c4eeaf
/registry/events/default/node-02.15b9e0af99ba73a2
/registry/events/default/node-02.15b9e0ca43c5e760
/registry/events/default/node-03.15b9e0ae9bdae96c
/registry/events/default/node-03.15b9e0aea880813c
/registry/events/default/node-03.15b9e0aea880ae05
/registry/events/default/node-03.15b9e0aea880c0de
/registry/events/default/node-03.15b9e0aeb13cfeef
/registry/events/default/node-03.15b9e0afcbcf299b
/registry/events/default/node-03.15b9e0b02f28fa3c
/registry/events/default/node-03.15b9e0cadf7dce89
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9ddb67e6ab700
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9e0af3bdb47fe
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9e0cbbbb7579d
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9e0cc279fbd33
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9e0cc34fb8de2
/registry/events/kube-system/coredns-5c98db65d4-5kjjv.15b9e0cc4994ad54
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9ddb6850e5ff1
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0aea988964f
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0cbbb3af928
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0cc2ffb9d11
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0cc3a4def6c
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0cc4bd20265
/registry/events/kube-system/coredns-5c98db65d4-88hkq.15b9e0cc6e488534
/registry/events/kube-system/kube-proxy-7642v.15b9e0ae1444b38c
/registry/events/kube-system/kube-proxy-7642v.15b9e0ae7ff6f434
/registry/events/kube-system/kube-proxy-7642v.15b9e0af631fa3d0
/registry/events/kube-system/kube-proxy-7642v.15b9e0af7632698a
/registry/events/kube-system/kube-proxy-7642v.15b9e0af85356aad
/registry/events/kube-system/kube-proxy-jsp4r.15b9e0aeadc2ce3a
/registry/events/kube-system/kube-proxy-jsp4r.15b9e0af27535c1b
/registry/events/kube-system/kube-proxy-jsp4r.15b9e0affc7fc79e
/registry/events/kube-system/kube-proxy-jsp4r.15b9e0b00a290340
/registry/events/kube-system/kube-proxy-jsp4r.15b9e0b01b0a4eef
/registry/events/kube-system/kube-proxy.15b9e0ae1333a730
/registry/events/kube-system/kube-proxy.15b9e0aeaad76df0
/registry/events/kube-system/weave-net-2hvbx.15b9e0c6e9b6c1de
/registry/events/kube-system/weave-net-2hvbx.15b9e0c71a365ad4
/registry/events/kube-system/weave-net-2hvbx.15b9e0c88a5af203
/registry/events/kube-system/weave-net-2hvbx.15b9e0c8a5998774
/registry/events/kube-system/weave-net-2hvbx.15b9e0c8b54252cb
/registry/events/kube-system/weave-net-2hvbx.15b9e0c8b5543df6
/registry/events/kube-system/weave-net-2hvbx.15b9e0c98384d3e1
/registry/events/kube-system/weave-net-2hvbx.15b9e0c9916478ce
/registry/events/kube-system/weave-net-2hvbx.15b9e0c9a090c521
/registry/events/kube-system/weave-net-5mrjl.15b9e0c6e9523ad2
/registry/events/kube-system/weave-net-5mrjl.15b9e0c7194191cb
/registry/events/kube-system/weave-net-5mrjl.15b9e0c89c46497c
/registry/events/kube-system/weave-net-5mrjl.15b9e0c8b335c817
/registry/events/kube-system/weave-net-5mrjl.15b9e0c8c714f12d
/registry/events/kube-system/weave-net-5mrjl.15b9e0c8c770ebdd
/registry/events/kube-system/weave-net-5mrjl.15b9e0c995196184
/registry/events/kube-system/weave-net-5mrjl.15b9e0c9a24d099d
/registry/events/kube-system/weave-net-5mrjl.15b9e0c9b2e0cdef
/registry/events/kube-system/weave-net-c76fx.15b9e0c6ec0133eb
/registry/events/kube-system/weave-net-c76fx.15b9e0c7255593bb
/registry/events/kube-system/weave-net-c76fx.15b9e0c8d4f52821
/registry/events/kube-system/weave-net-c76fx.15b9e0c90ebfeb95
/registry/events/kube-system/weave-net-c76fx.15b9e0c922410c3a
/registry/events/kube-system/weave-net-c76fx.15b9e0c922580ded
/registry/events/kube-system/weave-net-c76fx.15b9e0c9f7834364
/registry/events/kube-system/weave-net-c76fx.15b9e0ca15411664
/registry/events/kube-system/weave-net-c76fx.15b9e0ca2d254f2c
/registry/events/kube-system/weave-net.15b9e0c6e7edf622
/registry/events/kube-system/weave-net.15b9e0c6e9c8d2c1
/registry/events/kube-system/weave-net.15b9e0c6ea880cd2
/registry/leases/kube-node-lease/node-01
/registry/leases/kube-node-lease/node-02
/registry/leases/kube-node-lease/node-03
/registry/masterleases/134.209.178.162
/registry/minions/node-01
/registry/minions/node-02
/registry/minions/node-03
/registry/namespaces/default
/registry/namespaces/kube-node-lease
/registry/namespaces/kube-public
/registry/namespaces/kube-system
/registry/pods/kube-system/coredns-5c98db65d4-5kjjv
/registry/pods/kube-system/coredns-5c98db65d4-88hkq
/registry/pods/kube-system/etcd-node-01
/registry/pods/kube-system/kube-apiserver-node-01
/registry/pods/kube-system/kube-controller-manager-node-01
/registry/pods/kube-system/kube-proxy-7642v
/registry/pods/kube-system/kube-proxy-jsp4r
/registry/pods/kube-system/kube-proxy-xj8qm
/registry/pods/kube-system/kube-scheduler-node-01
/registry/pods/kube-system/weave-net-2hvbx
/registry/pods/kube-system/weave-net-5mrjl
/registry/pods/kube-system/weave-net-c76fx
/registry/priorityclasses/system-cluster-critical
/registry/priorityclasses/system-node-critical
/registry/ranges/serviceips
/registry/ranges/servicenodeports
/registry/replicasets/kube-system/coredns-5c98db65d4
/registry/rolebindings/kube-public/kubeadm:bootstrap-signer-clusterinfo
/registry/rolebindings/kube-public/system:controller:bootstrap-signer
/registry/rolebindings/kube-system/kube-proxy
/registry/rolebindings/kube-system/kubeadm:kubelet-config-1.15
/registry/rolebindings/kube-system/kubeadm:nodes-kubeadm-config
/registry/rolebindings/kube-system/system::extension-apiserver-authentication-reader
/registry/rolebindings/kube-system/system::leader-locking-kube-controller-manager
/registry/rolebindings/kube-system/system::leader-locking-kube-scheduler
/registry/rolebindings/kube-system/system:controller:bootstrap-signer
/registry/rolebindings/kube-system/system:controller:cloud-provider
/registry/rolebindings/kube-system/system:controller:token-cleaner
/registry/rolebindings/kube-system/weave-net
/registry/roles/kube-public/kubeadm:bootstrap-signer-clusterinfo
/registry/roles/kube-public/system:controller:bootstrap-signer
/registry/roles/kube-system/extension-apiserver-authentication-reader
/registry/roles/kube-system/kube-proxy
/registry/roles/kube-system/kubeadm:kubelet-config-1.15
/registry/roles/kube-system/kubeadm:nodes-kubeadm-config
/registry/roles/kube-system/system::leader-locking-kube-controller-manager
/registry/roles/kube-system/system::leader-locking-kube-scheduler
/registry/roles/kube-system/system:controller:bootstrap-signer
/registry/roles/kube-system/system:controller:cloud-provider
/registry/roles/kube-system/system:controller:token-cleaner
/registry/roles/kube-system/weave-net
/registry/secrets/default/default-token-nz988
/registry/secrets/kube-node-lease/default-token-4w7tf
/registry/secrets/kube-public/default-token-pzhnr
/registry/secrets/kube-system/attachdetach-controller-token-69mzv
/registry/secrets/kube-system/bootstrap-signer-token-584pq
/registry/secrets/kube-system/bootstrap-token-w1d2kx
/registry/secrets/kube-system/certificate-controller-token-rff4s
/registry/secrets/kube-system/clusterrole-aggregation-controller-token-6hks4
/registry/secrets/kube-system/coredns-token-b2874
/registry/secrets/kube-system/cronjob-controller-token-55pgx
/registry/secrets/kube-system/daemon-set-controller-token-nhcdf
/registry/secrets/kube-system/default-token-f5kl4
/registry/secrets/kube-system/deployment-controller-token-lm58l
/registry/secrets/kube-system/disruption-controller-token-4tw6s
/registry/secrets/kube-system/endpoint-controller-token-qdh8q
/registry/secrets/kube-system/expand-controller-token-6stw5
/registry/secrets/kube-system/generic-garbage-collector-token-hqfqx
/registry/secrets/kube-system/horizontal-pod-autoscaler-token-h6czj
/registry/secrets/kube-system/job-controller-token-nmw8f
/registry/secrets/kube-system/kube-proxy-token-zcrj8

上面显示了一共342个定义在配置文件中的key，包含所有在集群里面的资源：

Nodes
Namespaces
ServiceAccounts
Roles,RolesBindings, ClusterRoles/ClusterRoleBindings
ConfigMaps
Secrets
Workloads: Deployments, DaemonSets, Pods, …
Cluster’s certificates
The resources within each apiVersion
The events that bring the cluster in the current state

如果我们选择以上其中一个key，我们可以得到具体与之关联的value通过以下命令

$ kubectl exec etcd-node-01 -n kube-system —- sh -c \
"ETCDCTL_API=3 etcdctl \
--endpoints $ADVERTISE_URL \
--cacert /etc/kubernetes/pki/etcd/ca.crt \
--key /etc/kubernetes/pki/etcd/server.key \
--cert /etc/kubernetes/pki/etcd/server.crt \
get \"KEY\" -w json"

举个例子，我们想获得/registry/deployments/kube-system/coredns这个key的value

如果我们将这个对应的value解码出来发现信息其实可读性不高，一些图表不能被正常解码显示，但当然，Kubernetes内部是知道如何正确处理他们的。

从这个结果上看，我们可以推断出这个key是用来存储credns这个Pod的规则以及部署状态的。

创建一个Pod

让我们一起来创建一个Pod，看看集群的状态修改了什么以及有什么新的key的增加

$ cat <<EoF | kubectl apply -f -
apiVersion: v1
kind: Pod
metadata:
  name: www
spec:
  containers:
  - name: nginx
    image: nginx:1.16-alpine
EoF

和之前一样的命令，我们获取所有的key并保存在etcd-kv-after-nginx-pod.json中。快速对比一下我们创建Pod之前的key列表和我们创建完wwwPod以后多了什么信息

> /registry/events/default/www.15b9e3051648764f
> /registry/events/default/www.15b9e3056b8ce3f0
> /registry/events/default/www.15b9e306918312ea
> /registry/events/default/www.15b9e306a32beb6d
> /registry/events/default/www.15b9e306b5892b60
> /registry/pods/default/www

5个event以及一个pod被创建了，并且看上去也十分合理。我们更深入地看一下，从解码event key所对应的value开始，按照时间排序，我们可以看到他们做了下面的事情：

成功指派default/www到node-02
拉取镜像nginx:1.16-alpine
成功拉取镜像nginx:1.16-alpine
创建nginxcontainer
启动这个container

这些事件可以通过下面命令查看

kubectl describe pod www

最后一个key，/registry/pods/default/www，提供了这个新创建的pod的所有信息

最后使用的配置
关联的token
状态
...

（依然解码出来可读性很糟糕。）

总结

这篇文章的目的不是深入etcd，而是解释了一小部分Kubernetes内部包含了什么信息以及这些信息是怎么被组织起来的。希望能填补你对Kubernetes的一小片空白。