Kubernetes监控——Heapster+InfluxDB+Grafana

Kubernetes & Docker监控之最后一章，前三篇如下：

1、kubernetes+docker监控之简介

2、 kubernetes+docker监控之Docker监控——cadvisor

3、 Docker监控——Cadvisor+InfluxDB+Grafana搭建过程

1）、heapster下载和启动：

（1）、heapster镜像下载：

docker pull index.tenxcloud.com/google_containers/heapster:v1.1.0

（2）、heapster单独作为容器启动：

docker run -d index.tenxcloud.com/google_containers/heapster:v1.1.0 "--source=kubernetes:http://192.168.16.100:8080?inClusterConfig=false&kubeletHttps=true&kubeletPort=10250&useServiceAccount=true&auth=" "--sink=influxdb:http://192.168.16.234:8086"

参数说明：

--source ：指定kube-apiserver，这里使用https，需要注意下

--sink ：指定influxdb，heapster自动创建数据库【k8s】

inClusterConfig ：Use kube config in service accounts associated with heapster's

namesapce. (default: true)

kubeletPort ：kubelet port to use (default: 10255)

kubeletHttps ： whether to use https to connect to kubelets (default: false)

apiVersion ： API version to use to talk to Kubernetes. Defaults to the version

in kubeConfig.

insecure ：whether to trust kubernetes certificates (default: false)

auth ：client auth file to use. Set auth if the service accounts are not usable.

useServiceAccount ：whether to use the service account token if one is mounted

at /var/run/secrets/kubernetes.io/serviceaccount/token (default: false)

heapster容器，启动过程：

[root@localhost kube-1.2]# docker run -d index.tenxcloud.com/google_containers/heapster:v1.1.0 "--source=kubernetes:http://192.168.16.100:8080?inClusterConfig=false&kubeletHttps=true&kubeletPort=10250&useServiceAccount=true&auth=" "--sink=influxdb:http://192.168.16.234:8086"

b86aaaaa88ccf648120611254bfeb0078a8b9dd853d4b5a2d6a0abb97743bf1d

[root@localhost kube-1.2]# docker logs b86aaaaa8

I0714 07:05:18.116787 1 heapster.go:65] /heapster --source=kubernetes:http://192.168.16.100:8080?inClusterConfig=false&kubeletHttps=true&kubeletPort=10250&useServiceAccount=true&auth= --sink=influxdb:http://192.168.16.234:8086

I0714 07:05:18.117158 1 heapster.go:66] Heapster version 1.1.0

I0714 07:05:18.239220 1 configs.go:60] Using Kubernetes client with master "http://192.168.16.100:8080" and version "v1"

I0714 07:05:18.239270 1 configs.go:61] Using kubelet port 10250

I0714 07:05:31.255132 1 influxdb.go:223] created influxdb sink with options: host:192.168.16.234:8086 user:root db:k8s

I0714 07:05:31.255170 1 heapster.go:92] Starting with InfluxDB Sink

I0714 07:05:31.255177 1 heapster.go:92] Starting with Metric Sink

I0714 07:05:31.271159 1 heapster.go:171] Starting heapster on port 8082

I0714 07:06:05.000403 1 manager.go:79] Scraping metrics start: 2016-07-14 07:05:00 +0000 UTC, end: 2016-07-14 07:06:00 +0000 UTC

E0714 07:06:05.000504 1 kubelet.go:270] No nodes received from APIserver.

I0714 07:06:05.000520 1 manager.go:152] ScrapeMetrics: time: 5.756µs size: 0

I0714 07:06:05.365486 1 influxdb.go:201] Created database "k8s" on influxDB server at "192.168.16.234:8086"

[root@localhost kube-1.2]#

[root@localhost kube-1.2]# docker ps |grep heapster

b86aaaaa88cc index.tenxcloud.com/google_containers/heapster:v1.1.0 "/heapster --source=k" 2 minutes ago Up 2 minutes furious_mahavira

[root@localhost kube-1.2]#

看到influxdb已经有k8s的数据库和表了，剩下的就是grafana搜索和显示了

[root@localhost kube-1.2]# influx

Visit https://enterprise.influxdata.com to register for updates, InfluxDB server management, and monitoring.

Connected to http://localhost:8086 version 0.9.6

InfluxDB shell 0.9.6

>

>

> show databases

name: databases

---------------

name

_internal

cadvisor

k8s

>

> use k8s

Using database k8s

>

> SHOW MEASUREMENTS

name: measurements

------------------

name

cpu/node_reservation

cpu/node_utilization

cpu/usage

cpu/usage_rate

filesystem/limit

filesystem/usage

memory/major_page_faults

memory/major_page_faults_rate

memory/node_reservation

memory/node_utilization

memory/page_faults

memory/page_faults_rate

memory/usage

memory/working_set

network/rx

network/rx_errors

network/rx_errors_rate

network/rx_rate

network/tx

network/tx_errors

network/tx_errors_rate

network/tx_rate

uptime

2）、InfluxDB安装与配置：

请参考之前Cadvisor+InfluxDB+Grafana的内容，http://my.oschina.net/fufangchun/blog/718382

3）、Grafana安装与配置：

（1）、Grafana安装：

请参考之前Cadvisor+InfluxDB+Grafana的内容，http://my.oschina.net/fufangchun/blog/718382

（2）、Grafana配置：

Ⅰ、配置influxdb数据源：

Ⅱ、新建展示页：

展示页，用来区分不同的influxdb数据源

Ⅲ、新建数据项：

根据特定的条件（sql），从influxdb查询相关的数据，并展示。

Grafana会自动显示k8s库的字段，可以根据需求进行选择，每个where后面，加上：【pod_name =~ /^\.*/】，按照pod名称分类

SELECT last(value) FROM k8s."default"."network/tx" WHERE pod_name =~ /^\.*/ AND time > now() - 30m GROUP BY time(10s), pod_name

Ⅳ、完整的grafana展示：

这里上传可能看不清，可以去百度盘下载，我这里是有些数据是分别做了，按照pod和namespace进行分组的，有多租户的情况，比较实用，可以根据情况，自行定义。

本文转移开源中国-Kubernetes监控——Heapster+InfluxDB+Grafana