[TOC]

0x00 前言简述

描述: 最近公司有业务需要搞一个自动化的 jmeter 分布式压力测试项目, 基于现有的 Kubernetes 、Jenkins、 Gitlab 技术栈环境, 本章实践了在 Kubernetes 集群中使用 helm 搭建 jmeter 分布式压力测试 Master 与 server 以及 Influxdb 时序数据库 和 Grafana 环境，在 Windows 中 安装 jmeter 编写带有线程组的压力测试任务，在线程组中利用变量参数传递线程数和循环此处，然后再添加后端监听器设置为influxdb, 针对压力测试数据进行采集, 导出压力测试jmx文件，上传到 Gitlab 创建的 jmeter 仓库中，此处我们已经将Jenkins + Gitlab自动化集成部署已经打通(可以参考下面文章进行相应的学习配置)，在上传到 jmeter 代码仓库后将会通过 webhook 请求 jenkins 流水线项目从而触发自动化部署操作，然后Jenkins会使用kubenetes插件在集群中创建一个jenkins slave Pod(插件的使用以及slave 镜像可以在下面实践中获得)，之后便会执行流水线中三个步骤即代码拉取、压力测试、结果展示，并且会将部署执行情况通过企业微信提供的webhook发送到指定运维群中，我们可以非常方便点击grafana中 Jmeter 展示数据的 dashboard 地址进行查看 Jmeter 相关压力测试指标数据等，然后针对压力测试数据进行应用和部署环境的调优。

技术栈介绍
基于公司现有Kubernetes 、Jenkins、 Gitlab 技术栈,实现在devops Jenkins CI/CD 自动化环境中使用 jmeter-distributed 针对部署的正式或者测试环境进行 jmeter 分布式压力测试，并且使用InfluxDB存储压力测试数据，和利用 Grafana进行压力测试数据的实时展示，在测试、运维、开发工作可以借此压力测试数据针对开发应用不断优化，保证上线后承载能力。

• Kubernetes : Google 在 2014 年开源了 Kubernetes 项目。 Kubernetes 建立在Google 大规模运行生产工作负载十几年经验，它是一个可移植、可扩展的开源平台,用于管理容器化的工作负载和服务,方便进行声明式配置和自动化。
• GitLab : 是一个用于仓库管理系统的开源项目，使用Git作为代码管理工具(类似于 Github)，并在此基础上搭建起来的Web服务, 你可以将其认为是在企业中私有化代码仓库。
• Jenkins : 一个开源自动化服务器，使世界各地的开发人员能够可靠地构建、测试和部署他们的软件，即 CI/CD 持续集成与交互。
• Grafana : 开放、可组合的可观察性和数据可视化平台，可视化来自普罗米修斯（Prometheus）、洛基（Loki）、Elasticsearch、InfluxDB、Postgres等多个来源的度量、日志和跟踪。
• Jmeter : 是 Apache 组织基于 Java 开发的压力测试工具，用于对软件做压力测试。
• InfluxDB : 是一个开源时间序列平台(数据库), 包括用于存储和查询数据、在后台处理数据以用于 ETL 或监控和警报目的、用户仪表板以及可视化和探索数据等的 API。

好的, 废话不多说实践才是王道。

本章完整原文地址:

• 企业运维实践-如何在K8S集群环境Gitlab+Jenkins+Jmeter+Grafana技术中实现自动化分布压力测试数据展示
• https://mp.weixin.qq.com/s/-Gbrtno8MOfqgCdgWxKE0w

---

0x01 安装配置

1.基础环境

环境说明

$ kubectl get node
  # NAME       STATUS   ROLES                  AGE    VERSION
  # weiyigeek-107   Ready    control-plane,master   186d   v1.23.1
  # weiyigeek-108   Ready    control-plane,master   185d   v1.23.1
  # weiyigeek-109   Ready    control-plane,master   185d   v1.23.1
  # weiyigeek-223   Ready    work                   185d   v1.23.1
  # weiyigeek-224   Ready    work                   185d   v1.23.1
  # weiyigeek-225   Ready    work                   186d   v1.23.1
  # weiyigeek-226   Ready    work                   25d    v1.23.1
  
# 其它辅助软件工具
Java     1.8.0_251
Jmeter   5.4.3
Jenkins  2.330
InfluxDB 1.8.10
Grafana  v9.0.2

在K8S集群中部署动态持久卷

描述: 此处使用nfs动态持久卷主要是用于PV/PVC来存放Pod中需要持久化存储的数据。

参考链接: https://blog.weiyigeek.top/2022/6-7-664.html#1-集群中基于nfs的provisioner的动态持卷环境部署

# (1) 挂载测试的nfs共享目录到本地/storage/dev中
sudo mkdir -vp /storage/dev
sudo tee -a /etc/fstab <<'EOF'
192.168.12.32:/k8sdevtest /storage/dev nfs defaults 0 0
EOF
sudo mount -a

# (2) 部署nfs-subdir-external-provisioner到storage名称空间下。
kubectl create ns storage
helm install nfs-devtest nfs-subdir-external-provisioner-dev/ --debug --namespace storage

# (3) 查看集群中的创建 storageclass nfs-devtest
kubectl get storageclasses.storage.k8s.io nfs-devtest
  # NAME                    PROVISIONER                                    RECLAIMPOLICY   VOLUMEBINDINGMODE   ALLOWVOLUMEEXPANSION   AGE
  # nfs-devtest (default)   cluster.local/nfs-devtest-subdir-provisioner   Delete          Immediate           true                   8m13s

在K8S集群中部署压力测试演示站点

# 1.此处，拉取我主页站点静态资源为例进行压力测试站点演示
mkdir -vp /nfs/app/web && git https://github.com/WeiyiGeek/WeiyiGeek.git

# 2.准备一个 Nginx 配置文件
tee /nfs/app/web/nginx.conf <<'EOF'
user  nginx;
worker_processes  auto;
error_log  /var/log/nginx/error.log notice;
pid        /var/run/nginx.pid;
events {
  worker_connections  1024;
}
http {
  include       /etc/nginx/mime.types;
  default_type  application/octet-stream;
  log_format  main  '$remote_addr - $remote_user [$time_local] "$request" $status $body_bytes_sent "$http_referer" "$http_user_agent" "$http_x_forwarded_for"';
  access_log  /var/log/nginx/access.log  main;
  sendfile        on;
  keepalive_timeout  65;
  gzip  on;
  server {
    listen       80;
    listen  [::]:80;
    server_name  localhost;
    #access_log  /var/log/nginx/host.access.log  main;
    location / {
      root   /usr/share/nginx/html;
      index  index.html index.htm;
    }
  }
}
EOF

kubectl create cm nginx-conf --from-file=nginx.conf --namespace devtest

# 3.使用StatefulSet资源控制器部署nginx为静态资源提供web访问。
tee /nfs/app/web/jmeter-test-html.yaml <<'EOF'
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: jmeter-test-html
  namespace: devtest
  labels:
    app: web-html
spec:
  replicas: 1
  selector:
    matchLabels:
      app: web-html
  serviceName: "jmeter-test-html"
  template:
    metadata:
      labels:
        app: web-html
    spec:
      # affinity:
      #   nodeAffinity:
      #     requiredDuringSchedulingIgnoredDuringExecution:
      #       nodeSelectorTerms:
      #       - matchExpressions:
      #         - key: node
      #           operator: In
      #           values:
      #           - work
      #   podAntiAffinity:
      #     preferredDuringSchedulingIgnoredDuringExecution:
      #     - podAffinityTerm:
      #         labelSelector:
      #           matchExpressions:
      #           - key: app
      #             operator: In
      #             values:
      #             - web-html
      #         topologyKey: kubernetes.io/hostname
      #       weight: 100
      volumes:
      - name: workdir
        emptyDir: {}
      - name: static-file
        hostPath:
          path: /nfs/app/web/WeiyiGeek
          type: DirectoryOrCreate
      - name: nginx-conf
        configMap:
          name: nginx-conf
          items:
          - key: nginx.conf
            path: nginx.conf
      - name: timezone
        hostPath:
          path: /usr/share/zoneinfo/Asia/Shanghai
      initContainers:
      - name: sysctl
        image: alpine:3.15.4
        imagePullPolicy: IfNotPresent
        command:
        - sh
        - -c
        - |
          mount -o remount rw /proc/sys
          sysctl -w net.core.somaxconn=65535
          sysctl -w net.ipv4.tcp_tw_reuse=1
          sysctl -w net.ipv4.ip_local_port_range="1024 65535"
          sysctl -w fs.file-max=1048576
          sysctl -w fs.inotify.max_user_instances=16384
          sysctl -w fs.inotify.max_user_watches=524288
          sysctl -w fs.inotify.max_queued_events=16384
        securityContext:
          privileged: true
      containers:
      - name: nginx
        image: nginx:1.21.6
        imagePullPolicy: IfNotPresent
        ports:
        - name: http
          protocol: TCP
          containerPort: 80
        volumeMounts:
        - name: static-file
          mountPath: /usr/share/nginx/html
        - name: nginx-conf
          mountPath: /etc/nginx/nginx.conf
          subPath: nginx.conf
---
apiVersion: v1
kind: Service
metadata:
  name: jmeter-test-html
  namespace: devtest
spec:
  type: nodePort
  ports:
    - port: 80
      targetPort: 80
      protocol: TCP
      nodePort: 32644
  selector:
    app: web-html
EOF
kubectl apply -f jmeter-test-html.yaml

# 4.验证部署的nginx压力测试站点
$ kubectl get svc -n devtest jmeter-test-html
  # NAME               TYPE       CLUSTER-IP    EXTERNAL-IP   PORT(S)        AGE
  # jmeter-test-html   NodePort   10.103.88.2   <none>        80:32644/TCP   37d
$ curl -s 10.103.88.2

温馨提示: 如有同道之人在进行实践时,请更改上述文件目录为您实际存放静态资源以及nginx配置的目录。

2.依赖环境

在 Windows 中安装 Apache jmeter 工具

描述: 此处针对于Windows中安装JAVA环境以及 Apache jmeter不在累述, 如果不会安装的同学可以参考 (1.使用Apache Jmeter对应用压力测试学习与实践)[https://blog.weiyigeek.top/2022/6-11-661.html#Windows] 这篇文章或者参考Apache jmeter的官网。

以二进制方式安装Helm部署工具

快速部署

ARCH=$(dpkg --print-architecture)
VERSION="3.8.2"
wget https://get.helm.sh/helm-v${VERSION}-linux-${ARCH}.tar.gz -O /tmp/helm-v${VERSION}-linux-${ARCH}.tar.gz
cd /tmp/ && tar -zxf helm-v${VERSION}-linux-${ARCH}.tar.gz
sudo cp linux-amd64/helm /usr/local/bin/

---

以 helm 方式安装 Grafana 9.x

快速部署

# 添加 chart 仓库与索引更新
$ helm repo add grafana https://grafana.github.io/helm-charts
$ helm repo update
$ helm search repo grafana 
  # grafana/grafana   6.32.3   9.0.3   The leading tool for querying and visualizing t...

# 获取指定 chart 版本的granfa相关部署参数
$ helm show values grafana/grafana --version 6.32.3  > granfa-values.yaml  

# 安装指定 chart 版本以及部署参数
$ helm install grafana grafana/grafana --version 6.32.3  --set nodeSelector="kubernetes\.io/hostname=node-223",service.type=NodePort,persistence.enabled=True,persistence.storageClassName="persistentVolumeClaim",persistence.size=5Gi,persistence.accessModes=ReadWriteOnce -n dashboard --debug --create-namespace

# 获取 grafana 初始化登陆密码
kubectl get secrets -n devops grafana -o yaml
echo "RlJLMkR1bmlpalY2YXRMcGtuU053NmhLanRibnhZZUZRY25tV0ZtSg==" | base64 -d

# 删除部署的grafana应用
$ helm delete grafana

---

0x02 Kubernetes + jmeter + InfluxDB 1.8.x + Grafana + Jenkins 进行自动化分布式压力测试及其结果展示

(1) 以 helm 方式安装分布式 jmeter 压力测试应用

快速部署

# 添加 chart 仓库与索引更新
$ helm repo add apphub https://apphub.aliyuncs.com/
$ helm repo update

# 搜索可用的 chart 仓库中软件版本
$ helm search repo jmeter
  # NAME                            CHART VERSION   APP VERSION     DESCRIPTION
  # apphub/distributed-jmeter       1.0.1           3.3             A Distributed JMeter Helm chart

# 获取指定 chart 版本的 distributed-jmeter 相关部署参数
$ helm show values apphub/distributed-jmeter --version 1.0.1 > distributed-jmeter.yaml
$ cat distributed-jmeter.yaml
# Default values for distributed-jmeter.
master:
  replicaCount: 1
server:
  replicaCount: 3
image:
  ## Specify an imagePullPolicy
  ## ref: http://kubernetes.io/docs/user-guide/images/#pre-pulling-images
  pullPolicy: IfNotPresent
  ## The repository and image
  ## ref: https://hub.docker.com/r/pedrocesarti/jmeter-docker/
  repository: "pedrocesarti/jmeter-docker"
  ## The tag for the image
  ## ref: https://hub.docker.com/r/pedrocesarti/jmeter-docker/tags/
  tag: 3.2

# 安装 distributed-jmeter 应用到 devtest 名称空间下
$ helm install jmeter apphub/distributed-jmeter -f distributed-jmeter.yaml --namespace devtest --create-namespace
# To get get a shell session on the master you only need to run:
  $ export MASTER_NAME=$(kubectl get pods -l app.kubernetes.io/component=master -o jsonpath='{.items[*].metadata.name}'  -n devtest )
  $ kubectl -n devtest exec -it $MASTER_NAME  -- /bin/bash
# To copy your test plans to the master pod
  $ kubectl -n devtest  cp sample.jmx $MASTER_NAME:/jmeter
# To run your test in all servers you need first a list of all servers IPs (comma-separated) and then you can run your test:
  $ export SERVER_IPS=$(kubectl get pods -l app.kubernetes.io/component=server -o jsonpath='{.items[*].status.podIP}'  -n devtest | tr ' ' ',')
  $ kubectl exec -it $MASTER_NAME -- jmeter -n -t /jmeter/sample.jmx -R $SERVER_IPS

# 查看 helm 部署资源
$ kubectl get deployments.apps,pod,svc -n devtest -l app.kubernetes.io/name=distributed-jmeter
  # NAME                                               READY   UP-TO-DATE   AVAILABLE   AGE
  # deployment.apps/jmeter-distributed-jmeter-master   1/1     1            1           117m
  # deployment.apps/jmeter-distributed-jmeter-server   3/3     3            3           117m

  # NAME                                                    READY   STATUS    RESTARTS   AGE
  # pod/jmeter-distributed-jmeter-master-5d5587ffb-8ftvk    1/1     Running   0          114m
  # pod/jmeter-distributed-jmeter-server-7d64577d56-9cdk5   1/1     Running   0          111m
  # pod/jmeter-distributed-jmeter-server-7d64577d56-bpf4q   1/1     Running   0          112m
  # pod/jmeter-distributed-jmeter-server-7d64577d56-sjsd2   1/1     Running   0          114m

  # NAME                                       TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)              AGE
  # service/jmeter-distributed-jmeter-server   ClusterIP   None         <none>        50000/TCP,1099/TCP   117m

手动测试部署环境
步骤 01.准备一个简单的 jmeter 测试任务，例如下图中Apache jmeter里定义的 jmeter-test.jmx 任务。

WeiyiGeek.jmeter-test.jmx

步骤 02.将jmeter中测试任务导出到jmeter-test.jmx文件中，并复制到 Master 节点容器中，然后执行分布式压力测试命令。

# 获取 Master 节点 以及分布式压力测试节点 IP 地址
export MASTER_NAME=$(kubectl get pods -l app.kubernetes.io/component=master -o jsonpath='{.items[*].metadata.name}'  -n devtest )
export SERVER_IPS=$(kubectl get pods -l app.kubernetes.io/component=server -o jsonpath='{.items[*].status.podIP}' -n devtest | tr ' ' ',')

# 手动拷贝 jmx 压力测试配置文件到Jmeter 主节点中
kubectl -n devtest cp ./jmeter-test.jmx  ${MASTER_NAME}:/jmeter/jmeter-test.jmx

# 单节点压力测试 指定 -Jgroup.threads 线程数 ，以及 -Jgroup.loops 线程循环次数。
kubectl -n devtest exec -it $MASTER_NAME -- jmeter -Jgroup.threads=5 -Jgroup.loops=20 -n -t /jmeter/jmeter-test.jmx -l jmeter-test.result -R $SERVER_IPS 

# 分布式压力测试 指定 -Ggroup.threads 每个节点线程数 ，以及 -Ggroup.loops 线程循环次数。（此处坑有点大）
kubectl -n devtest exec -it $MASTER_NAME -- jmeter -Ggroup.threads=20 -Ggroup.loops=20 -n -t /jmeter/jmeter-test.jmx -l jmeter-test.result -R $SERVER_IPS 
  # Creating summariser <summary>
  # Created the tree successfully using /jmeter/jmeter-test.jmx
  # Configuring remote engine: 10.66.24.209
  # Configuring remote engine: 10.66.183.112
  # Configuring remote engine: 10.66.182.204
  # Starting remote engines
  # Starting the test @ Mon Jun 13 04:49:42 UTC 2022 (1655095782003)
  # Remote engines have been started
  # Waiting for possible Shutdown/StopTestNow/Heapdump message on port 4445
  # summary =   1200 in 00:00:02 =  569.0/s Avg:     4 Min:     0 Max:   685 Err:     0 (0.00%)
  # Tidying up remote @ Mon Jun 13 04:49:46 UTC 2022 (1655095786021)
  # ... end of run

---

(2) 以资源清单方式安装配置 InfluxDB 1.8 时序数据库

快速安装

步骤 01.准备influxdb使用的配置文件并存放到configMap资源控制器下。

mkdir -vp /storage/dev/webapp/database/jmeter-influxdb
tee /storage/dev/webapp/database/jmeter-influxdb/influxdb.conf <<'EOF'
[meta]
  dir = "/var/lib/influxdb/meta"
  retention-autocreate = true
  logging-enabled = true

[data]
  dir = "/var/lib/influxdb/data"
  index-version = "inmem"
  wal-dir = "/var/lib/influxdb/wal"
  wal-fsync-delay = "0s"

[coordinator]
  write-timeout = "10s"
  max-concurrent-queries = 0
  query-timeout = "0s"
  log-queries-after = "0s"
  max-select-point = 0
  max-select-series = 0
  max-select-buckets = 0

[retention]
  enabled = true
  check-interval = "30m0s"

[monitor]
  store-enabled = true
  store-database = "_internal"
  store-interval = "10s"

[subscriber]
  enabled = true
  http-timeout = "30s"
  insecure-skip-verify = false
  ca-certs = ""
  write-concurrency = 40
  write-buffer-size = 1000

[shard-precreation]
  enabled = true
  check-interval = "10m0s"
  advance-period = "30m0s"

[http]
  enabled = true
  bind-address = ":8086"

[logging]
  format = "auto"
  level = "info"
  suppress-logo = false

[[graphite]]
  enabled = true
  bind-address = ":2003" 
  database = "jmeter"
  retention-policy = ""
  protocol = "tcp"
  batch-size = 5000
  batch-pending = 10
  batch-timeout = "1s"
  consistency-level = "one"
  separator = "."
  udp-read-buffer = 0

[continuous_queries]
  log-enabled = true
  enabled = true
  query-stats-enabled = false
  run-interval = "1s"
EOF

$ cd /storage/dev/webapp/database/jmeter-influxdb
$ kubectl create cm influxdb-config --from-file=influxdb.conf --namespace database
  # configmap/influxdb1-config created

步骤 02.准备使用Deployment资源控制器部署jmeter-influxDB1.x相关的资源清单如下：

tee influxdb-Deployment.yaml <<'EOF'
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: jmeter-influxdb-pvc
  namespace: database
  labels:
    app: jmeter-influxdb
spec:
  storageClassName: nfs-devtest
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 2Gi
---
apiVersion: v1
kind: Service
metadata:
  name: jmeter-influxdb
  namespace: database
  labels:
    app: jmeter-influxdb
spec:
  ports:
  - name: api
    nodePort: 32086
    port: 8086
    protocol: TCP
    targetPort: 8086
  - name: graphite
    nodePort: 32003
    port: 2003
    protocol: TCP
    targetPort: 2003
  selector:
    app: jmeter-influxdb
  sessionAffinity: None
  type: NodePort
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: jmeter-influxdb
  namespace: database
  labels:
    app: jmeter-influxdb
spec:
  replicas: 1
  selector:
    matchLabels:
      app: jmeter-influxdb
  template:
    metadata:
      labels:
        app: jmeter-influxdb
    spec:
      containers:
      - image: influxdb:1.8.10
        imagePullPolicy: IfNotPresent
        name: influxdb
        ports:
        - containerPort: 8086
          name: api
          protocol: TCP
        - containerPort: 2003
          name: graphite
          protocol: TCP
        volumeMounts:
        - mountPath: /etc/influxdb/influxdb.conf
          name: config
          subPath: influxdb.conf
        - mountPath: /var/lib/influxdb
          name: data
      restartPolicy: Always
      terminationGracePeriodSeconds: 30
      volumes:
      - name: config
        configMap:
          defaultMode: 420
          name: influxdb-config
          items:
          - key: influxdb.conf
            path: influxdb.conf
      - name: data
        persistentVolumeClaim:
          claimName: jmeter-influxdb-pvc
      # - name: data-local
      #  hostPath:
      #    path: /storage/database/influxdb/data
      #    type: DirectoryOrCreate     
EOF

步骤 03.使用kubectl工具按照influxdb-Deployment.yaml资源清单中的内容进行安装部署, 并且查看其部署情况。

$ kubectl apply -f influxdb-Deployment.yaml
  # persistentvolumeclaim/jmeter-influxdb-pvc created
  # service/jmeter-influxdb created
  # deployment.apps/jmeter-influxdb created

$ kubectl get pod,svc,pvc -n database -l app=jmeter-influxdb
  # NAME                                   READY   STATUS    RESTARTS   AGE
  # pod/jmeter-influxdb-7867969fbd-ggt4m   1/1     Running   0          56s
  
  # NAME                      TYPE       CLUSTER-IP     EXTERNAL-IP   PORT(S)                         AGE
  # service/jmeter-influxdb   NodePort   10.106.17.52   <none>        8086:32086/TCP,2003:32003/TCP   5m12s
  
  # NAME                                        STATUS   VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
  # persistentvolumeclaim/jmeter-influxdb-pvc   Bound    pvc-d260839d-86cc-4f34-976f-50563ce28084   2Gi        RWO            nfs-devtest    6m9s
  
# 补充说明:
# - Pod 端口转发
kubectl port-forward --namespace database $(kubectl get pods --namespace database -l app=jmeter-influxdb -o jsonpath='{ .items[0].metadata.name }') 8086:8086

# - Pod Shell终端
kubectl exec -i -t --namespace database $(kubectl get pods --namespace database -l app=jmeter-influxdb -o jsonpath='{.items[0].metadata.name}') /bin/bash

# - Pod 日志查看
kubectl logs -f --namespace database $(kubectl get pods --namespace database -l app=jmeter-influxdb -o jsonpath='{ .items[0].metadata.name }')

步骤 04.进入 jmeter-influxdb pod 的 shell 终端中创建 jmeter 所使用的用户及其数据库。

$ kubectl exec -i -t --namespace database $(kubectl get pods --namespace database -l app=jmeter-influxdb -o jsonpath='{.items[0].metadata.name}') /bin/bash

root@influxdb-f54d77bc6-cp46r:/# influx
Connected to http://localhost:8086 version 1.8.10
InfluxDB shell version: 1.8.10
# 查看交互式帮助
> help 
Usage:
  connect <host:port>   connects to another node specified by host:port
  auth                  prompts for username and password
  pretty                toggles pretty print for the json format
  chunked               turns on chunked responses from server
  chunk size <size>     sets the size of the chunked responses.  Set to 0 to reset to the default chunked size
  use <db_name>         sets current database
  format <format>       specifies the format of the server responses: json, csv, or column
  precision <format>    specifies the format of the timestamp: rfc3339, h, m, s, ms, u or ns
  consistency <level>   sets write consistency level: any, one, quorum, or all
  history               displays command history
  settings              outputs the current settings for the shell
  clear                 clears settings such as database or retention policy.  run 'clear' for help
  exit/quit/ctrl+d      quits the influx shell

  show databases        show database names
  show series           show series information
  show measurements     show measurement information
  show tag keys         show tag key information
  show field keys       show field key information

# influx 客户端默认设置查看
> settings
  Setting           Value
  --------          --------
  URL               http://localhost:8086
  Username
  Database
  RetentionPolicy
  Pretty            false
  Format            column
  Write Consistency all
  Chunked           true
  Chunk Size        0

# 创建jmeter用户及其认证密码以及创建数据库jmeter
CREATE USER "jmeter" WITH PASSWORD 'weiyigeek.top' WITH ALL PRIVILEGES
CREATE DATABASE jmeter
SHOW DATABASES
USE jmeter
SHOW MEASUREMENTS
exit

---

(3) 在 jmeter 中添加 Backend Listener 配置 influxDB 相关信息

操作步骤
步骤 01.在(任务线程组⚙)右键选中 -> Add -> Listener -> Backend Listener > 点击后将会显示步骤02界面。

WeiyiGeek.add Backend Listener

步骤 02.在 Name 处输入 jmeter-influxDB 名称 , 在 Backend Listener implementation 中下拉选中( org.apache.jmeter.visualizers.backend.influxdb.influxdbBackendListenerClient ) , 其配置如下所示:

influxdbMetricsSender	org.apache.jmeter.visualizers.backend.influxdb.HttpMetricsSender
influxdbUrl	http://192.168.12.107:32086/write?db=jmeter
application	weiyigeek-index
measurement	jmeter
summaryOnly	true
samplersRegex	.*
percentiles	50;90;95;99
testTitle	jmeter-index
eventTags

WeiyiGeek.Configure Backend Listener

温馨提示: influxdbUrl 此处采用了 nodePort 暴露的端口

步骤 03.在线程组中添加一个HTTP-Request请求，将上面我们搭建演示压力测试的站点信息进行填入。

Name: myblog-test
Comments: 备注可以自定义
Protocol: http
Server NAME or IP: 192.168.12.107
Port Number: 32644
HTTP Request: GET
Path: /index.html
Content encoding: utf-8

WeiyiGeek.Configure HTTP Request

温馨提示：此处压力测试地址实际就是(http://192.168.12.107:32644/index.html).

步骤 04.配置完成后点击菜单栏中 《绿色start》 的按钮开始压力测试任务，请求 100 次 (此处缺省 5 线程 循环 20 ), 然后可以通过kubectl logs -f --tail 100 -n devtest jmeter-test-html-0命令查看到jmeter请求该路径的访问日志。

WeiyiGeek.

步骤 05.查看后端监听器配置influxdb数据中采集到的数据。

kubectl exec -i -t --namespace database $(kubectl get pods --namespace database -l app=jmeter-influxdb -o jsonpath='{.items[0].metadata.name}') /bin/bash -- influx

Connected to http://localhost:8086 version 1.8.10
InfluxDB shell version: 1.8.10
> use jmeter
Using database jmeter
> SHOW MEASUREMENTS
name: measurements
name
----
events
jmeter
> select * from jmeter; # 查看 jmeter 表中压力测试的数据
name: jmeter
time                application     avg                count countError endedT hit max maxAT meanAT min minAT pct50.0 pct90.0 pct95.0            pct99.0            rb      sb    startedT statut transaction
----                -----------     ---                ----- ---------- ------ --- --- ----- ------ --- ----- ------- ------- -------            -------            --      --    -------- ------ -----------
1658221734017000000 weiyigeek-index                                     0              0     0          0                                                                         0               internal
1658221735666000000 weiyigeek-index 2.09               100   0                 100 24               1         2       2       3.8999999999999773 23.799999999999898 1447500 16400          all    all
1658221735667000000 weiyigeek-index                                     5              1     0          0                                                                         5               internal

> select * from events; # 查看 jmeter 压力测试事件信息
name: events
time                application     text                 title
----                -----------     ----                 -----
1658221734002000000 weiyigeek-index jmeter-index started ApacheJMeter
1658221735666000000 weiyigeek-index jmeter-index ended   ApacheJMeter

(4) 在 grafana 中展示 jmeter 压力测试相关指标数据。

• 步骤 01.登陆Grafana之后需要新增数据源（Add your first data source）,此处当然选择InfluxDB时间序列数据库源。

WeiyiGeek.添加 InfluxDB 数据源

• 步骤 02.选择后将会出现如下界面, 需要按照实际情况进行填写, 其中最重要的是influxdb相关连接数据库认证信息（jmeter、jmeter、weiyigeek.top）, 填写无误后点击 Save&Test, 如测试成功将会显示Data source is working。

温馨提示: 此处 influxdb url 填入的是 (http://jmeter-influxdb.database.svc:8086) 即服务名称加SVC端口，当然你也可以填写节点IP+nodePort端口即(http://192.168.12.107:32086/)

• 步骤 03.然后可以点击步骤02页面上的Explore，执行如下sql：SELECT * FROM "jmeter" limit 3进行查询到jmeter压力测试的数据。

WeiyiGeek.InfluxDB Explore

• 步骤 04.访问grafana模板商城 (https://grafana.com/grafana/dashboards/?plcmt=footer&search=jmeter), 找寻显示 jmeter 相应的 dashboards 模板, 此处以模板ID 5496 为例 (https://grafana.com/grafana/dashboards/5496), 在 grafana 中 import 输入该ID -> 点击 load -> 随后设置 dashboard 名称 -> DBNAME 选择前面添加的 InfluxDB 数据源 -> Measurement name 设置为 jmeter -> 后端发送间隔时间设置为 5 即可 -> 最后点 Import 导入该 Dashboard 到 grafana 中.

WeiyiGeek.Apache JMeter Dashboard using Core InfluxdbBackendListenerClient

• 步骤 05.导入后我们便可在 Grafana 首页中查看到该 Dashboard, 我们点击进入查看该展板显示的 Jmeter 压力测试的数据。

WeiyiGeek.Apache JMeter Dashboard

(5) 使用 kubernetes + Gitlab + Jenkins + Grafana 实现自动化分布式压力测试数据展示实践

实践流程:
步骤 01.首先在私有的 Gitlab 上创建一个Jmeter，此处我已经在Gitlab中创建好了jmeter项目, 你可以在创建后将前面准备的jmeter 测试任务 jmeter-test.jmx 文件push 到代码仓库中。

WeiyiGeek.Jmeter 仓库

温馨提示: Gitlab 的安装此处不进行展开讲解, 如有需要安装实践的童鞋可以参考我的这篇文章 [GitLab企业级私有代码仓库安装与基础使用] (https://blog.weiyigeek.top/2019/7-22-98.html) 或者 gitlab 官方安装文档 (https://about.gitlab.com/install/)。

步骤 02.在 Jenkins 中创建一个名为 Jmeter 流水线的任务，然后再进行如下配置:

(1) 勾选 GitHub 项目 -> 输入Jmeter项目地址 (http://gitlab.weiyigeek.top/devops/jmeter.git)

(2) 在 GitLab Connection 下拉框中选择 Gitlab-Auth 认证的票据 (其生成可以参考下面温馨提示中的文章)

(3) 构建触发器选项卡中勾选 Enabled GitLab triggers.
Build when a change is pushed to GitLab. GitLab webhook URL: http://192.168.12.226:30001/project/jmeter

温馨提示: Jenkins 的相关安装部署请参考我的博客文章 《https://blog.weiyigeek.top/tags/Jenkins/》

步骤 03.在 Jmeter 项目仓库中添加拥有Reporter权限的自动化CI/CD用户, 此处添加的 @devops 用户，为了能自动触发Jenkins进行部署以及压力测试，我们需要再设置 -> webhooks -> 添加 jenkins 任务 url 以及 认证 Secret Token

URL: http://192.168.12.226:30001/project/jmeter
Secret Token: 11c513a49620v573d275357836cc3e1c4f

温馨提示: 上述 Secret Token 的创建 以及Gitlab私有仓库生成项目API Access Token可以参考我的如下两篇文章。

• Jenkins与Gitlab-集成配置与实践 https://blog.weiyigeek.top/2020/12-28-513.html#Gitlab-集成配置与实践

• Gitlab 自动触发构建之 Pipeline Script from SCM https://blog.weiyigeek.top/2020/12-30-523.html#2-Gitlab-自动触发构建之-Pipeline-Script-from-SCM

步骤 04.在 jenkins Jmeter 任务中流水线的脚本如下:

// [ Groovy 全局变量定义]
// - GIT项目地址
// def GITLAB = "http://gitlab.weiyigeek.top/devops/jmeter.git" 
def PRJECT = "ssh://git@gitlab.weiyigeek.top/devops/jmeter.git"
def PUBKEY = "5e2f46d9-6725-4399-847f-dafb3f29d0ce"

// - 企业微信 - WebHookURL
def QY_WECHAT_WEBHOOK_KEY= "ee1165c3-56c1-4bc1-8423-ecc5ccda9278"
def QY_WECHAT_WEBHOOK = "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${QY_WECHAT_WEBHOOK_KEY}"

// - 正式环境与测试环境参数预设
def ENV_TEST() {
  def config = [:]
  config.K8S_API_SERVER = "https://slb-vip.k8s:16443"
  config.K8S_IP_SERVER = "192.168.12.110"
  config.K8S_CREDENTIALSID = "8cf24244-3e26-41a9-b7fe-3eb20778c33a"
  config.CI_NAME = "${env.JOB_NAME}"
  config.CI_NAMESPACE = "devtest"
  config.CI_ENVIRONMENT = "test"
  config.CI_COMMAND = '["bash"]'
  config.CI_STORAGE_NAME = "nfs-devtest"
  config.INGRESS_DOMAIN = "testapp.weiyigeek.top"
  config.HARBOR_URL = "harbor.weiyigeek.top/devops"
  config.HARBOR_AUTH = "d0ce1239-c4bf-41ed-a4c6-660ab70d9b47"
  return config
}

def ENV_PROD() {
  def config = [:]
  config.K8S_API_SERVER = "https://apiserver.cluster.weiyigeek:6443"
  config.K8S_IP_SERVER = "192.168.12.102"
  config.K8S_CREDENTIALSID = "k8s-cluster.weiyigeek-devops"
  config.CI_NAME = "${env.JOB_NAME}"
  config.CI_NAMESPACE = "devtest"
  config.CI_ENVIRONMENT = "prod"
  config.CI_COMMAND = '["bash"]'
  config.CI_STORAGE_NAME = "nfs-storage"
  config.INGRESS_DOMAIN = "test.app.weiyigeek.top"
  config.HARBOR_URL = "harbor.weiyigeek.top/devops"
  config.HARBOR_AUTH = "d0ce1239-c4bf-41ed-a4c6-660ab70d9b47"
  return config
}

// 任务名称、任务工作空间POD名称
def JOB_NAME = "${env.JOB_NAME}-${env.BUILD_NUMBER}"
def JOB_WORKSPACE = "${env.WORKSPACE}"

// 自定义变量
// String JmeterNAME = null
// String JmeterENV = null
// String JmeterFILE = null
// Integer JmeterGthreads = null
// Integer JmeterGloops = null

// [ Groovy 全局函数定义]
// 获取真实项目路径机器构建的包信息函数
def GET_REAL_PROJECT(){
  def PROJECT = [:]
  PROJECT.commitmsg=sh label: 'git_commitmsg',returnStdout: true, script: """
   git show --oneline --ignore-all-space --text | head -n 1 |tr -d '\\n'
  """
  // 验证与提取commit信息 45cd579 jmeter cloud zycx.jmx 100 1000
  if( PROJECT.commitmsg != null ) {
    def arr = PROJECT.commitmsg.split(" ") as List
    print arr
    PROJECT.JmeterNAME = arr[1]
    PROJECT.JmeterENV = arr[2]
    PROJECT.JmeterFILE = arr[3]
    PROJECT.JmeterGthreads = arr[4].toInteger()
    PROJECT.JmeterGloops = arr[5].toInteger()
  } else {
    unstable '提取git commit信息失败'
  }

  // 获取 jmx 任务文件路径
  PROJECT.JmeterFilePath=sh label: 'JmeterFilePath',returnStdout: true, script: """
    find ${env.WORKSPACE} -name ${PROJECT.JmeterFILE} | head -n 1 | tr -d '\n'
  """
  print PROJECT.JmeterFilePath

  return PROJECT
}


// [Jenkins Pipeline 流水线 开始]
pipeline {
  // 流水线运行的主机绑定,此处利用动态的K8s节点进行。
  agent {
    kubernetes {
      cloud 'kubernetes'
      namespace 'devops'
      inheritFrom 'jenkins-slave'
      showRawYaml 'false'
      workingDir '/home/jenkins/agent'
      yaml """\
apiVersion:
kind: Pod
metadata:
  labels:
    app: 'jenkins-jnlp'
    job: ${JOB_NAME}
spec:
  serviceAccountName: 'jenkins-sa'
  automountServiceAccountToken: false
  securityContext:
    runAsUser: 1000   # default UID of jenkins user in agent image
  containers:
  - name: 'jnlp'
    image: 'harbor.weiyigeek.top/devops/alpine-jenkins-jnlp:v2.330'
    imagePullPolicy: 'IfNotPresent'
    command: ["/bin/sh"]
    args: ["-c","/usr/local/bin/jenkins-agent.sh && cat"]
    tty: true
    env:
      - name: JAVA_OPTS
        value: '-Xms512m -Xmx1g -Xss1m'
    resources:
      limits: {}
      requests:
        memory: '512Mi'
        cpu: '500m'
    volumeMounts:
    - name: 'docker-socket'
      mountPath: '/var/run/docker.sock'
  volumes:
  - name: docker-socket
    hostPath: 
      path: /var/run/docker.sock
    """
    }
  }
  
  // 全局选项预定义设定
  options {
   gitLabConnection('Private-Gitlab')
   gitlabBuilds(builds: ['代码拉取', '压力测试','结果展示'])
   timeout(time: 30, unit: 'MINUTES') 
  }
  
  // 自定义环境变量, 通过 env.变量名访问 
  environment {
    // 代码仓库地址与认证地址
    GITLAB_URL = "${PRJECT}"
    GITLAB_PUB = "${PUBKEY}"

    // 代码质量检测项目名称与反馈超时时间
    SONARQUBE_PROJECTKEY = "${JOB_NAME}";
    SONARQUBE_TIMEOUT = '60';
  }

  // 自定义选择参数，在 sh 中可通过变量名访问，而在 script pipeline 脚本中通过 params.参数名称 访问 
  parameters {
    gitParameter branch: '', branchFilter: 'origin/(.*)', defaultValue: 'origin/master', description: '查看构建部署可用的Tag或Branch名称?', name: 'TagBranchName', quickFilterEnabled: false, selectedValue: 'NONE', sortMode: 'DESCENDING_SMART', tagFilter: '*', type: 'GitParameterDefinition'
    string(name: 'RELEASE_VERSION', defaultValue: "master", description: 'Message: 请选择构建部署的Tag或Branch名称?', trim: 'True')
    choice(name: 'PREJECT_ENVIRONMENT', choices: ['Test', 'Test'], description: 'Message: 选择项目部署环境?')
    choice(name: 'PREJECT_OPERATION', choices: ['None', 'deploy', 'rollback', 'redeploy'], description: 'Message: 选择项目操作方式?')
    choice(name: 'IS_IMAGEBUILD', choices: ['True','False'], description: 'Message: 是否进行镜像构建操作?')
    choice(name: 'IS_RELEASE', choices: ['False','True'], description: 'Message: 是否进行编译成品发布?')
    choice(name: 'IS_SONARQUBE', choices: ['False','True'], description: 'Message: 是否进行代码质量检测?')
  }

  // 触发与Gitlab 流水线同步更新
  triggers {
    gitlab(triggerOnPush: true, triggerOnMergeRequest: true, branchFilterType: 'All')
  } 

  // 主要阶段以及子阶段流程
  stages {
    // [ 阶段1.项目代码拉取 ]
    stage ('代码拉取') {
      steps {
        // 1.超时时间设置5分钟
        timeout(time: 5, unit: 'MINUTES') {
          script {
            try {
              checkout([$class: 'GitSCM', branches: [[name: "${params.RELEASE_VERSION}"]], doGenerateSubmoduleConfigurations: false, extensions: [], submoduleCfg: [], userRemoteConfigs: [[credentialsId: "${env.GITLAB_PUB}", url: "${env.GITLAB_URL}"]]])
              updateGitlabCommitStatus name: '代码拉取', state: 'success'
            } catch(Exception err) {
              updateGitlabCommitStatus name: '代码拉取', state: 'failed'
              // * 显示错误信息但还是会进行下一阶段操作
              echo err.toString()    /* hudson.AbortException: Couldn't find any revision to build. Verify the repository and branch configuration for this job. */
              unstable '拉取代码失败'
              error "[-Error] : 代码拉取失败\n [-Msg] : ${err.getMessage()} " /* 失败立即停止 */
            } 
        
            // 2.构建信息输出 （注意必须放入到 script 代码块中）
            if ( "${env.gitlabActionType}" != "null" ) {
              echo "任务名称: ${JOB_NAME}, 自动触发操作: ${env.gitlabActionType}, 项目名称: ${env.gitlabSourceRepoName}, 项目地址: ${env.GITLAB_URL},项目分支：${env.gitlabBranch} - ${env.gitlabSourceBranch}"
            } else {
              echo "任务名称: ${JOB_NAME}, 手动触发操作: ${params.PREJECT_OPERATION}, 项目环境: ${env.PREJECT_ENVIRONMENT}"
            }

            // 3.利用 GET_REAL_PROJECT 函数获取项目相关信息
            project = GET_REAL_PROJECT()

            // 4.验证部署环境进行预设配置环境参数值
            if ( "${project.JmeterENV}" == "cloud" ) {
              config = ENV_TEST()
              print "项目环境: ${project.JmeterENV}"
            } else {
              config = ENV_TEST()
              print "项目环境: ${project.JmeterENV}"
            }

            // 5.通过企业微信进行构建通知
            if ( "${env.gitlabActionType}" != "null" ) {
              sh label: 'auto message notification', script: """\
              curl ${QY_WECHAT_WEBHOOK} \
              -H 'Content-Type: application/json' \
              -d '{
                  "msgtype": "markdown",
                  "markdown": {
                    "content":"Jenkins-消息通知<font color=\\"info\\">【${JOB_NAME}】</font>任务开始, 自动触发! \\n
                    >项目信息: <font color=\\"warning\\">${project.JmeterNAME} - ${env.gitlabBranch} - ${env.gitlabMergeRequestLastCommit}</font>
                    >提交信息: <font color=\\"warning\\">${project.commitmsg}</font>
                    >测试环境: ${project.JmeterENV}
                    >测试任务: ${project.JmeterFILE}
                    >线程数量: ${project.JmeterGthreads}
                    >循环次数: ${project.JmeterGloops}
                    >[查看当前任务流水线](${env.BUILD_URL})"
                  }
              }'
              """
            } else {
              sh label: 'manual message notification', script: """\
              curl ${QY_WECHAT_WEBHOOK} \
              -H 'Content-Type: application/json' \
              -d '{
                  "msgtype": "markdown",
                  "markdown": {
                    "content":"Jenkins-消息通知<font color=\\"info\\">【${JOB_NAME}】</font>任务开始, 手工触发! \\n
                    >项目信息: <font color=\\"warning\\">${project.JmeterNAME} - ${params.TagBranchName}</font>
                    >提交信息: <font color=\\"warning\\">${project.commitmsg}</font>
                    >测试环境: ${project.JmeterENV}
                    >测试任务: ${project.JmeterFILE}
                    >线程数量: ${project.JmeterGthreads}
                    >循环次数: ${project.JmeterGloops}
                    >[查看当前任务流水线](${env.BUILD_URL})"
                  }
              }'
              """
            }
          }
        }
     }
   }

    stage ("压力测试") {
      when {
        expression { return params.PREJECT_OPERATION == "None"}
      }
      steps {
        script {
            try {
              // 使用K8S集群票据进行操作远程集群部署资源
              withKubeConfig(caCertificate: '', clusterName: '', contextName: '', credentialsId: "${config.K8S_CREDENTIALSID}", namespace: '', serverUrl: "${config.K8S_API_SERVER}") {
                String[] K8S_API_DOMAIN;
                K8S_API_DOMAIN = "${config.K8S_API_SERVER}".split('//');
                K8S_API_DOMAIN = "${K8S_API_DOMAIN[1]}".split(':');
                print "K8S_API_DOMAIN ->" + K8S_API_DOMAIN[0];

                // 配置修改/etc/hosts文件以将Kubernetes API地址进行绑定,并判断jmeter Server Pod 状态是否正常
                res_status=sh label:'jmeter',returnStdout: true, script: """ 
                  echo "${config.K8S_IP_SERVER} ${K8S_API_DOMAIN[0]}" | sudo tee -a /etc/hosts  >> /dev/null \
                  && kubectl get pod -n ${config.CI_NAMESPACE} -l app.kubernetes.io/instance=jmeter | grep "jmeter" | wc -l | head -n 1 |tr -d '\\n'
                """

                // 如果为0则代表无该资源进行部署 
                if ( "${res_status}" != "0" ) {
                  print "jmeter-distributed (master/server) = ${res_status}"
                  // Master 节点名称
                  def MASTER_NAME=sh label:'MASTER_NAME',returnStdout: true, script: """
                    kubectl get pods -l app.kubernetes.io/component=master -o jsonpath='{.items[*].metadata.name}' -n ${config.CI_NAMESPACE} | tr -d '\\n'
                  """
                  // Server 节点IP
                  def SERVER_IPS=sh label:'SERVER_IPS',returnStdout: true, script: """
                    kubectl get pods -l app.kubernetes.io/component=server -o jsonpath='{.items[*].status.podIP}' -n ${config.CI_NAMESPACE} | tr ' ' ',' |tr -d '\\n'
                  """
                  print "Name: ${MASTER_NAME}\nServer_IPS: ${SERVER_IPS}"
      
                  // # 手动拷贝 jmx 压力测试配置文件到Jmeter 主节点中
                  // # 分布式压力测试 指定 -Ggroup.threads 每个节点线程数 ，以及 -Ggroup.loops 线程循环次数。（此处坑有点大）
                  sh label:'Jmeter-distributed', script: """
                    kubectl -n ${config.CI_NAMESPACE} cp ${project.JmeterFilePath} ${MASTER_NAME}:/jmeter/${project.JmeterFILE} && \
                    kubectl -n ${config.CI_NAMESPACE} exec -it $MASTER_NAME -- jmeter -Ggroup.threads=${project.JmeterGthreads} -Ggroup.loops=${project.JmeterGloops} -n -t /jmeter/${project.JmeterFILE} -l ${project.JmeterFILE}.result -R $SERVER_IPS 
                  """
                } else {
                  print "jmeter-distributed (master/server) = ${res_status}"
                  error "[-Error] : jmeter-distributed 分布式压力测试master、server环境未启动!\n "
                }
              }
              updateGitlabCommitStatus name: '压力测试', state: 'success'
            } catch(Exception err) {
              echo err.toString()    /* hudson.AbortException: Couldn't find any revision to build. Verify the repository and branch configuration for this job. */
              unstable '压力测试失败'
              updateGitlabCommitStatus name: '压力测试', state: 'failed'
              error "[-Error] : 压力测试失败\n [-Msg] : ${err.getMessage()} "
            }
        }
      }
    }

    stage('结果展示') {
      when {
        expression { return params.PREJECT_OPERATION == "None"}
      }
      steps {
        script {
          try {
              sh label:'grafana-jmeter-dashboard',script:"""
                curl -s ${QY_WECHAT_WEBHOOK} \
                -H 'Content-Type: application/json' \
                -d '{
                    "msgtype": "markdown",
                    "markdown": {
                      "content":"Jenkins-消息通知<font color=\\"info\\">【${JOB_NAME}】</font> - ${project.JmeterFILE} 任务压力测试完毕 !^_^\\n
                      >图形访问地址: <font color=\\"warning\\">[https://devops.weiyigeek.top/grafana/d/p6hnPpg4z/apache-jmeter-dashboard?from=now-5m&to=now](https://devops.weiyigeek.top/grafana/d/p6hnPpg4z/apache-jmeter-dashboard?from=now-5m&to=now)</font>
                      >[查看当前任务流水线](${env.BUILD_URL})"
                    }
                }'
              """
              updateGitlabCommitStatus name: '结果展示', state: 'success'
          } catch (Exception err) {
              updateGitlabCommitStatus name: '结果展示', state: 'failed'
              echo err.toString()    /* hudson.AbortException: Couldn't find any revision to build. Verify the repository and branch configuration for this job. */
          }
        }
      }
    }
  }

  // (12) POST阶段当所有任务执行后触发进行工作空间的清理以及消息通知;
  post {
    always {
      deleteDir()  /* clean up our workspace */
    }
    success {
      echo "Pipeline ${env.JOB_NAME} 项目成功 ^_^"
      qyWechatNotification failNotify: true, webhookUrl: "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${QY_WECHAT_WEBHOOK_KEY}"
    }
    unstable {
      echo "Pipeline ${env.JOB_NAME} 项目不稳定 :/"
      qyWechatNotification failNotify: true, webhookUrl: "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${QY_WECHAT_WEBHOOK_KEY}"
    }
    failure {
      echo "Pipeline ${env.JOB_NAME} 项目失败 :("
      qyWechatNotification failNotify: true, webhookUrl: "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${QY_WECHAT_WEBHOOK_KEY}"
    }
    changed {
      echo "Pipeline ${env.JOB_NAME} 项目改变 .-."
      qyWechatNotification failNotify: true, webhookUrl: "https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${QY_WECHAT_WEBHOOK_KEY}"
    }
  }
}

温馨提示: 上述使用 Jenkins 自定义 jnlp-Slave 镜像 weiyigeek/alpine-jenkins-jnlp，如需自行构建可以参考如下地址https://hub.docker.com/r/weiyigeek/alpine-jenkins-jnlp

温馨提示: 此处在 jenkins 中使用 kubernetes-plugins 插件以便在 kubernetes Cloud 集群中动态生成 Slave Pod ， 参考地址: https://blog.weiyigeek.top/2020/12-30-531.html#3-集群动态创建-Agent-节点-Slave-节点。

温馨提示: 上述 webhook 消息通知是采用了企业微信群机器人 qy-wechat-notification-plugin 插件 https://blog.weiyigeek.top/2020/12-31-537.html#4-qy-wechat-notification-plugin，当然你也可以将其换为阿里钉钉通知，主要看你企业中是使用哪一个作为你的通知使用。

步骤 05.在前面的环境都准备好的情况下，我们便可以上传jmeter任务jmx格式的文件到gitlab仓库中，然后配置的webhook将会自动触发Jenkins任务从而进行自动化压力测试，此处我们先来看自动触发。

1.我们需要先创建一个 jmeter 带线程组的任务，在Thread Properties中分别在Number of Threads 、 Loop Count 配置为 ${__P(group.threads,50)} 和 ${__P(group.loops,200)}

2.然后在 jmeter 带线程组的任务, 右键新增一个 Backend Listner 其关键点如下所示:

# influxdbMetricsSender 一般默认即可
influxdbMetricsSender: org.apache.jmeter.visualizers.backend.influxdb.HttpMetricsSender
# 关键点: 搭建的influxdb 1.x连接字符串
influxdbUrl: http://192.168.12.107:32086/write?db=jmeter
# 关键点: 应用名称后续在grafana中需要手动选择指定应用名称的压力测试结果
application: index
# 关键点: jmeter 压力测试数据存放的表名称
measurement	jmeter
# 仅仅收集摘要信息
summaryOnly	true
# 采样器Regex正则表达式
samplersRegex	.*
# 统计的百分位
percentiles	50;90;95;99
# 测试标题
testTitle	title-index
# 事件标签
eventTags	tag-index

WeiyiGeek.Backend Listner

步骤 06.此处我使用 git 将 jmeter 项目克隆到本地中, 然后将上面的jmeter任务进行导出，并且复制 jmeter-test.jmx 到本地E:\githubProject\jmeter 目录之中，通过 git 三步法将更改提交到Gitlab仓库中。

git clone ssh://git@gitlab.weiyigeek.top:2222/devops/jmeter.git
cd E:\githubProject\jmeter
copy C:\Users\WeiyiGeek\Downloads\jmeter-test.jmx E:\githubProject\jmeter

$ git add .
$ git commit -m "jmeter weiyigeek jmeter-test.jmx 50 100"
$ git push

温馨提示：此处我已经将ssh公钥添加到gitlab中了, 后续拉取和推送代码到私有仓库更加方便。

步骤 07.在 Jenkins 中可以查看到 Started by GitLab push by WeiyiGeek 自动触发字样, 并且在 blueocean 插件界面下可以看到流水线执行的情况。

+ kubectl -n devtest cp /home/jenkins/agent/workspace/jmeter/jmeter-test.jmx jmeter-distributed-jmeter-master-5d5587ffb-b5nf6:/jmeter/jmeter-test.jmx
+ kubectl -n devtest exec -it jmeter-distributed-jmeter-master-5d5587ffb-b5nf6 -- jmeter '-Ggroup.threads=50' '-Ggroup.loops=100' -n -t /jmeter/jmeter-test.jmx -l jmeter-test.jmx.result -R 10.66.24.215,10.66.24.209,10.66.0.197

WeiyiGeek.Jenkins pipeline & blueocean

步骤 08.上述 Jenkins pipeline 流水线执行完毕之后，可以从企业微信中查看到 jmeter 任务相关通知

WeiyiGeek.qywechat notification

步骤 09.点击企业微信中展示压力测试数据的 Grafana dashboard 地址，查看jmeter压力测试的相关数据。

WeiyiGeek.Grafana & jmeter

步骤 10.上述提供的Jenkins流水线中除了通过Gitlab代码仓库更新自动触发外，我们还可以进行手动触发进行jmeter自动化压力测试，我们在Jenkins Dashboard 中点击 jmeter -> Build With Parameters -> 此处我只有一个默认分支所以，其参数缺省即可(请注意脚本中的ENV_TEST()函数) -> 点击开始构建, 完成后将如下图所示。

WeiyiGeek.手动触发Jmeter压力测试

至此在在Kubernetes集群中进行jmeter + InfluxDB + gitlab + Jenkins进行自动化jmeter分布式压力测试实践完毕!.

---

0x03 在Kubernetes集群(jmeter + InfluxDB 2.x + Grafana)中进行自动化分布式压力测试及结果展示

描述: 如果要使用最新 jmeter dashboard 模板，我们需要将 InfluxDB 版本切换为 2.x 并且分布式压力测试 jmeter 程序需要运行在 JDK 11 以上。

温馨提示: Java SE Development Kit 11.0.15.1 (Java SE 订阅者将至少在 2026 年 9 月之前收到 JDK 11 更新), 下载地址 https://www.oracle.com/java/technologies/downloads/#java11

systeminfo
  # OS 名称: Microsoft Windows 10 专业版
  # OS 版本: 10.0.19044 暂缺 Build 19044
  
java --version
  # java 11.0.15.1 2022-04-22 LTS
  # Java(TM) SE Runtime Environment 18.9 (build 11.0.15.1+2-LTS-10)
  # Java HotSpot(TM) 64-Bit Server VM 18.9 (build 11.0.15.1+2-LTS-10, mixed mode)

jmeter.bat 
  # Debug Output: "11.0.15.1"

(1) 以 helm 方式安装InfluxDB 2.x 时序数据库

快速安装

# - chart 仓库添加与索引更新
helm repo add bitnami https://charts.bitnami.com/bitnami 
helm repo update && helm search repo influxdb
  # NAME                    CHART VERSION   APP VERSION     DESCRIPTION
  # bitnami/influxdb        5.3.1           2.2.0           InfluxDB(TM) is an open source time-series data

# - chart 模板参数配置
helm show values bitnami/influxdb > influxdb.yml

# - 部署 influxdb2 
helm install jmeter-influxdb2 bitnami/influxdb -f influxdb.yml --version 5.3.1 -n devtest --debug --create-namespace
kubectl get pod -n devtest jmeter-influxdb2-5574c9c4c9-bzx4l jmeter-
  # NAME                                READY   STATUS    RESTARTS       AGE
  # jmeter-influxdb2-5574c9c4c9-bzx4l   1/1     Running   5 (148m ago)   153m

# - 进入 influxdb2 Pod 终端
kubectl exec -it -n devtest jmeter-influxdb2-5574c9c4c9-bzx4l -- bash
I have no name!@jmeter-influxdb2-5574c9c4c9-bzx4l:/$ env | egrep "^INFLUXDB_"
INFLUXDB_ADMIN_USER_TOKEN=9ONq6UdzvEtMRny40uUN
INFLUXDB_ADMIN_ORG=primary
INFLUXDB_CREATE_USER_TOKEN=no
INFLUXDB_ADMIN_USER=admin
INFLUXDB_HTTP_AUTH_ENABLED=true
INFLUXDB_ADMIN_BUCKET=primary
INFLUXDB_ADMIN_USER_PASSWORD=fL2JnkqExH

# - 创建 jmeter 的组织
influx org create -n jmeter -d "jmeter organization" -t 9ONq6UdzvEtMRny40uUN
  # ID                      Name
  # 65386f1047a4a0ef        jmeter

# - 创建 jmeter 组织下名为 jmeter 桶
influx bucket create -n jmeter -r 0 -o jmeter -t 9ONq6UdzvEtMRny40uUN
  # ID                      Name    Retention       Shard group duration    Organization ID         Schema Type
  # ba47a04e75fb2e05        jmeter  infinite        168h0m0s                65386f1047a4a0ef        implicit

# - 创建 jmeter 组织下名为 jmeter 用户
influx user create --name jmeter --password weiyigeek.top --org jmeter -t 9ONq6UdzvEtMRny40uUN
  # ID                      Name
  # 0982d4da8fe3b000        jmeter

# - 创建生成 jmeter 用户认证Token及其权限,该权限只在自己的组织中拥有权限。
influx auth create --user jmeter --org jmeter --all-access -t 9ONq6UdzvEtMRny40uUN
  # ID                      Description     Token                                                                                  User Name        User ID                 Permissions
  # 0982d54a42a3b000                        eoAyDV97es2jh0oC2oF2AWiMXupn9Hswq_axKUe8TKqvk7zCcEGeoQYUGA8WJS-tC4dzPg4Qaz01uEizmniWMw==jmeter          0982d4da8fe3b000        [read:orgs/65386f1047a4a0ef/authorizations write:orgs/65386f1047a4a0ef/authorizations read:orgs/65386f1047a4a0ef/buckets write:orgs/65386f1047a4a0ef/buckets read:orgs/65386f1047a4a0ef/dashboards write:orgs/65386f1047a4a0ef/dashboards read:/orgs/65386f1047a4a0ef read:orgs/65386f1047a4a0ef/sources write:orgs/65386f1047a4a0ef/sources read:orgs/65386f1047a4a0ef/tasks write:orgs/65386f1047a4a0ef/tasks read:orgs/65386f1047a4a0ef/telegrafs write:orgs/65386f1047a4a0ef/telegrafs read:/users/0982d4da8fe3b000 write:/users/0982d4da8fe3b000 read:orgs/65386f1047a4a0ef/variables write:orgs/65386f1047a4a0ef/variables read:orgs/65386f1047a4a0ef/scrapers write:orgs/65386f1047a4a0ef/scrapers read:orgs/65386f1047a4a0ef/secrets write:orgs/65386f1047a4a0ef/secrets read:orgs/65386f1047a4a0ef/labels write:orgs/65386f1047a4a0ef/labels read:orgs/65386f1047a4a0ef/views write:orgs/65386f1047a4a0ef/views read:orgs/65386f1047a4a0ef/documents write:orgs/65386f1047a4a0ef/documents read:orgs/65386f1047a4a0ef/notificationRules write:orgs/65386f1047a4a0ef/notificationRules read:orgs/65386f1047a4a0ef/notificationEndpoints write:orgs/65386f1047a4a0ef/notificationEndpoints read:orgs/65386f1047a4a0ef/checks write:orgs/65386f1047a4a0ef/checks read:orgs/65386f1047a4a0ef/dbrp write:orgs/65386f1047a4a0ef/dbrp read:orgs/65386f1047a4a0ef/notebooks write:orgs/65386f1047a4a0ef/notebooks read:orgs/65386f1047a4a0ef/annotations write:orgs/65386f1047a4a0ef/annotations read:orgs/65386f1047a4a0ef/remotes write:orgs/65386f1047a4a0ef/remotes read:orgs/65386f1047a4a0ef/replications write:orgs/65386f1047a4a0ef/replications]

# - 创建 dbrp 映射以便通过InfluxQL进行数据查询
influx v1 dbrp create --db jmeter_db --rp jmeter_rp --bucket-id ba47a04e75fb2e05 --org jmeter --default -t 9ONq6UdzvEtMRny40uUN
  # ID                      Database        Bucket ID               Retention Policy        Default Organization ID
  # 0982d6738223b000        jmeter_db       ba47a04e75fb2e05        jmeter_rp               true    65386f1047a4a0ef

# - 手动转发 influxdb 数据库端口到宿主机的 30086 端口上
kubectl port-forward --namespace devtest  --address 0.0.0.0 $(kubectl get pods --namespace devtest -l app.kubernetes.io/name=influxdb -o jsonpath='{ .items[0].metadata.name }') 30086:8086

0x0n 附录参考

Jmeter Backend Listener 组件介绍

• https://jmeter.apache.org/usermanual/component_reference.html#Backend_Listener
• https://jmeter.apache.org/usermanual/realtime-results.html

Jmeter 插件安装

描述: 存放插件jar目录的为D:\Tools\apache-jmeter-5.4.3\lib\ext,我们可以将jar文件将其放在/lib/ext中，重启Jmeter就可以了。

Prometheus Listener for Jmeter (为Prometheus导出JMeter指标), 利用它就能将 JMeter 变成一个 Data Exporter 然后在使用 Prometheus 进行抓取。
This JMeter plugin is highly configurable listener (and config element) to allow users define they’re own metrics (names, types etc.) and expose them through a Prometheus /metrics API to be scraped by a Prometheus server。

• https://github.com/johrstrom/jmeter-prometheus-plugin/releases/download/0.6.0/jmeter-prometheus-plugin-0.6.0.jar

• https://github.com/mderevyankoaqa/jmeter-influxdb2-listener-plugin/