StatefulSet 这个对象是专门用来部署用状态应用的，可以为 Pod 提供稳定的身份标识，包括 hostname、启动顺序、DNS 名称等。
下面以在 Kubernetes1.6 版本中部署 zookeeper 和 kafka 为例讲解 StatefulSet 的使用，其中 kafka 依赖于 zookeeper。
Dockerfile 和配置文件见 zookeeper 和 kafka。
注：所有的镜像基于 CentOS 系统的 JDK 制作，为我的私人镜像，外部无法访问，yaml 中没有配置持久化存储。
部署 Zookeeper
Dockerfile 中从远程获取 zookeeper 的安装文件，然后在定义了三个脚本：
- zkGenConfig.sh：生成 zookeeper 配置文件
- zkMetrics.sh：获取 zookeeper 的 metrics
- zkOk.sh：用来做 ReadinessProb
我们在来看下这三个脚本的执行结果。
zkMetrics.sh 脚本实际上执行的是下面的命令：
$ echo mntr | nc localhost $ZK_CLIENT_PORT >& 1
zk_version 3.4.6-1569965, built on 02/20/2014 09:09 GMT
zk_avg_latency 0
zk_max_latency 5
zk_min_latency 0
zk_packets_received 427879
zk_packets_sent 427890
zk_num_alive_connections 3
zk_outstanding_requests 0
zk_server_state leader
zk_znode_count 18
zk_watch_count 3
zk_ephemerals_count 4
zk_approximate_data_size 613
zk_open_file_descriptor_count 29
zk_max_file_descriptor_count 1048576
zk_followers 1
zk_synced_followers 1
zk_pending_syncs 0
zkOk.sh 脚本实际上执行的是下面的命令：
$ echo ruok | nc 127.0.0.1 $ZK_CLIENT_PORT
imok
zookeeper.yaml
下面是启动三个 zookeeper 实例的 YAML 配置文件：
---
apiVersion: v1
kind: Service
metadata:
name: zk-svc
labels:
app: zk-svc
spec:
ports:
- port: 2888
name: server
- port: 3888
name: leader-election
clusterIP: None
selector:
app: zk
---
apiVersion: v1
kind: ConfigMap
metadata:
name: zk-cm
data:
jvm.heap: "1G"
tick: "2000"
init: "10"
sync: "5"
client.cnxns: "60"
snap.retain: "3"
purge.interval: "0"
---
apiVersion: policy/v1beta1
kind: PodDisruptionBudget
metadata:
name: zk-pdb
spec:
selector:
matchLabels:
app: zk
minAvailable: 2
---
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
name: zk
spec:
serviceName: zk-svc
replicas: 3
template:
metadata:
labels:
app: zk
spec:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: "app"
operator: In
values:
- zk
topologyKey: "kubernetes.io/hostname"
containers:
- name: k8szk
imagePullPolicy: Always
image: harbor-001.jimmysong.io/library/zookeeper:3.4.6
resources:
requests:
memory: "2Gi"
cpu: "500m"
ports:
- containerPort: 2181
name: client
- containerPort: 2888
name: server
- containerPort: 3888
name: leader-election
env:
- name : ZK_REPLICAS
value: "3"
- name : ZK_HEAP_SIZE
valueFrom:
configMapKeyRef:
name: zk-cm
key: jvm.heap
- name : ZK_TICK_TIME
valueFrom:
configMapKeyRef:
name: zk-cm
key: tick
- name : ZK_INIT_LIMIT
valueFrom:
configMapKeyRef:
name: zk-cm
key: init
- name : ZK_SYNC_LIMIT
valueFrom:
configMapKeyRef:
name: zk-cm
key: tick
- name : ZK_MAX_CLIENT_CNXNS
valueFrom:
configMapKeyRef:
name: zk-cm
key: client.cnxns
- name: ZK_SNAP_RETAIN_COUNT
valueFrom:
configMapKeyRef:
name: zk-cm
key: snap.retain
- name: ZK_PURGE_INTERVAL
valueFrom:
configMapKeyRef:
name: zk-cm
key: purge.interval
- name: ZK_CLIENT_PORT
value: "2181"
- name: ZK_SERVER_PORT
value: "2888"
- name: ZK_ELECTION_PORT
value: "3888"
command:
- sh
- -c
- zkGenConfig.sh && zkServer.sh start-foreground
readinessProbe:
exec:
command:
- "zkOk.sh"
initialDelaySeconds: 10
timeoutSeconds: 5
livenessProbe:
exec:
command:
- "zkOk.sh"
initialDelaySeconds: 10
timeoutSeconds: 5
securityContext:
runAsUser: 1000
fsGroup: 1000
我们再主要下上面那三个脚本的用途。
部署 kafka
Kafka 的 docker 镜像制作跟 zookeeper 类似，都是从远程下载安装包后，解压安装。
与 zookeeper 不同的是，只要一个脚本，但是又依赖于我们上一步安装的 zookeeper，kafkaGenConfig.sh 用来生成 kafka 的配置文件。
我们来看下这个脚本。
#!/bin/bash
HOST=`hostname -s`
if [[ $HOST =~ (.*)-([0-9]+)$ ]]; then
NAME=${BASH_REMATCH[1]}
ORD=${BASH_REMATCH[2]}
else
echo "Failed to extract ordinal from hostname $HOST"
exit 1
fi
MY_ID=$((ORD+1))
sed -i s"/broker.id=0/broker.id=$MY_ID/g" /opt/kafka/config/server.properties
sed -i s'/zookeeper.connect=localhost:2181/zookeeper.connect=zk-0.zk-svc.brand.svc:2181,zk-1.zk-svc.brand.svc:2181,zk-2.zk-svc.brand.svc:2181/g' /opt/kafka/config/server.properties
该脚本根据 statefulset 生成的 pod 的 hostname 的后半截数字部分作为 broker ID，同时再替换 zookeeper 的地址。
Kafka.yaml
下面是创建 3 个 kafka 实例的 YAML 配置。
---
apiVersion: v1
kind: Service
metadata:
name: kafka-svc
labels:
app: kafka
spec:
ports:
- port: 9093
name: server
clusterIP: None
selector:
app: kafka
---
apiVersion: policy/v1beta1
kind: PodDisruptionBudget
metadata:
name: kafka-pdb
spec:
selector:
matchLabels:
app: kafka
minAvailable: 2
---
apiVersion: apps/v1beta1
kind: StatefulSet
metadata:
name: kafka
spec:
serviceName: kafka-svc
replicas: 3
template:
metadata:
labels:
app: kafka
spec:
affinity:
podAntiAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
- labelSelector:
matchExpressions:
- key: "app"
operator: In
values:
- kafka
topologyKey: "kubernetes.io/hostname"
podAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 1
podAffinityTerm:
labelSelector:
matchExpressions:
- key: "app"
operator: In
values:
- zk
topologyKey: "kubernetes.io/hostname"
terminationGracePeriodSeconds: 300
containers:
- name: k8skafka
imagePullPolicy: Always
image: harbor-001.jimmysong.io/library/kafka:2.10-0.8.2.1
resources:
requests:
memory: "1Gi"
cpu: 500m
env:
- name: KF_REPLICAS
value: "3"
ports:
- containerPort: 9093
name: server
command:
- /bin/bash
- -c
- "/opt/kafka/bin/kafkaGenConfig.sh && /opt/kafka/bin/kafka-server-start.sh /opt/kafka/config/server.properties"
env:
- name: KAFKA_HEAP_OPTS
value : "-Xmx512M -Xms512M"
- name: KAFKA_OPTS
value: "-Dlogging.level=DEBUG"
readinessProbe:
tcpSocket:
port: 9092
initialDelaySeconds: 15
timeoutSeconds: 1