PMM 對MYSQL 的監控配製

 

系統選擇： centos 7.2

    關閉防火牆：

    systemctl stop firewalld.service

    systemctl disable firewalld.service

    起用轉發：

    # vi /usr/lib/sysctl.d/00-system.conf

    添加以下代碼：

    net.ipv4.ip_forward=1

    重啓network服務

    # systemctl restart network

    查看是否修改爲功

    # sysctl net.ipv4.ip_forward

    若是返回爲「net.ipv4.ip_forward = 1」則表示成功了

   

pmm-server 安裝：

 　　pmm官方文檔:

　　　　Pmm 組成: https://www.percona.com/doc/percona-monitoring-and-management/architecture.html

　　　　PMM 安裝: https://www.percona.com/doc/percona-monitoring-and-management/deploy/server/docker.html#run-server-docker

 　　　相關開源項目：
　　　　　　https://www.sqlite.org/atomiccommit.html
　　　　　　https://github.com/prometheus

    1.安裝DOCKER，RPM包的名字是docker-io

     yum install docker -y

 

   2.啓動docker

     service docker start

     ps -ef |grep docker

 

   3.建立容器的數據磁盤

     docker create \

     -v /opt/prometheus/data \

     -v /opt/consul-data \

     -v /var/lib/mysql \

     -v /var/lib/grafana \                    

     --name pmm-data \

     docker.io/percona/pmm-server /bin/true   

     //默認本地不存pmm-server，會從官網下載 ,可是會很慢

   

    容器卷在宿主機的目錄位置：

    docker inspect pmm-data

 

        "Mounts": [
            {
                "Name": "134874af6736c9065c504d084eb0b754507da23a212cd90524d35463dcff8c54",
                "Source": "/var/lib/docker/volumes/134874af6736c9065c504d084eb0b754507da23a212cd90524d35463dcff8c54/_data",
                "Destination": "/var/lib/mysql",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "3d03268806f49143f8051aefc8ebef047ceeeeea24c5fe2828e06e0b08b36f6e",
                "Source": "/var/lib/docker/volumes/3d03268806f49143f8051aefc8ebef047ceeeeea24c5fe2828e06e0b08b36f6e/_data",
                "Destination": "/opt/consul-data",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "47dde5e315f8123018a485d8a922d060761a508b9958bc5e8c38f4b934152321",
                "Source": "/var/lib/docker/volumes/47dde5e315f8123018a485d8a922d060761a508b9958bc5e8c38f4b934152321/_data",
                "Destination": "/opt/prometheus/data",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "4faad60a6619be2f2aa29c34dd8630427cd70810d849a21bc26733b3970dfb6b",
                "Source": "/var/lib/docker/volumes/4faad60a6619be2f2aa29c34dd8630427cd70810d849a21bc26733b3970dfb6b/_data",
                "Destination": "/var/lib/grafana",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            }
        ],

   

  4.建立容器的運行實例：

    docker run -d \

    -p 80:80 \

    --volumes-from pmm-data \

    --name pmm-server \

    --restart always \

    percona/pmm-server:1.2.0

   

 

    容器卷在宿主機的目錄位置：

 

     docker inspect pmm-server

     "Mounts": [
            {
                "Name": "47dde5e315f8123018a485d8a922d060761a508b9958bc5e8c38f4b934152321",
                "Source": "/var/lib/docker/volumes/47dde5e315f8123018a485d8a922d060761a508b9958bc5e8c38f4b934152321/_data",
                "Destination": "/opt/prometheus/data",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "4faad60a6619be2f2aa29c34dd8630427cd70810d849a21bc26733b3970dfb6b",
                "Source": "/var/lib/docker/volumes/4faad60a6619be2f2aa29c34dd8630427cd70810d849a21bc26733b3970dfb6b/_data",
                "Destination": "/var/lib/grafana",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "134874af6736c9065c504d084eb0b754507da23a212cd90524d35463dcff8c54",
                "Source": "/var/lib/docker/volumes/134874af6736c9065c504d084eb0b754507da23a212cd90524d35463dcff8c54/_data",
                "Destination": "/var/lib/mysql",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            },
            {
                "Name": "3d03268806f49143f8051aefc8ebef047ceeeeea24c5fe2828e06e0b08b36f6e",
                "Source": "/var/lib/docker/volumes/3d03268806f49143f8051aefc8ebef047ceeeeea24c5fe2828e06e0b08b36f6e/_data",
                "Destination": "/opt/consul-data",
                "Driver": "local",
                "Mode": "",
                "RW": true,
                "Propagation": ""
            }
        ],

 

    更改面板數據sqlite庫存儲類型爲MYSQL(容器自帶)

    [root@workstation2017 ~]# docker exec -i -t pmm-server /bin/bash

    [root@86b5c5955521 opt]# vi /etc/grafana/grafana.ini

     查/sqlite 改成MYSQL 存儲

     type=mysql

     host=127.0.0.1:3306

     name=grafana

     user=root

     password =

     在容器進入mysql客戶端,建立數據庫grafana,

     create database grafana;

    

     生效：

     docker restart pmm-server

    // 默認面板監控數據數據保在  /var/lib/grafana/grafana.db  

 

PMM client 安裝:

    [root@workstation2017 ~]# service mysql start

    Redirecting to /bin/systemctl start  mysql.service

    下載與PMM SERVER 一致的 PMM client

    https://www.percona.com/downloads/pmm-client/

    

    配製：

    [root@workstation2017 ~]# pmm-admin config --server 192.168.79.220

    OK, PMM server is alive.

    PMM Server      | 192.168.79.220

    Client Name     | workstation2017

    Client Address  | 192.168.79.220

 

     開始監控：

    pmm-admin add mysql --query-source perfschema  --user root --password xx

     

    WEB訪問頁面並配製相關值：

    http://192.168.79.220/graph

 

    data sources--->edit data source

       config:

       　　 填入：

        　　name:prometheus

        　　type:prometheus

       http settings:加入

        　　http://127.0.0.1:9090/prometheus

    dashboards-->import

    點選導入import json文件

   

操做目的：監控指定的Metrics，減小系統的統計負載，並採用MYSQL做爲存儲後端  

詳細操做以下:

json文件容器導出到宿主機的目錄上,並在WEB 面板中導入

[root@f3e8318c8d7c dashboards]# pwd

/usr/share/percona-dashboards/dashboards

[root@f3e8318c8d7c dashboards]# ls

Amazon_RDS_OS_Metrics.json    MongoDB_Overview.json               MySQL_Performance_Schema.json      Prometheus.json

Cross_Server_Graphs.json      MongoDB_ReplSet.json                MySQL_Query_Response_Time.json     ProxySQL_Overview.json

Disk_Performance.json         MongoDB_RocksDB.json                MySQL_Replication.json             Summary_Dashboard.json

Disk_Space.json               MongoDB_WiredTiger.json             MySQL_Table_Statistics.json        System_Overview.json

MariaDB.json                  MySQL_InnoDB_Metrics.json           MySQL_TokuDB_Metrics.json          Trends_Dashboard.json

MongoDB_Cluster_Summary.json  MySQL_InnoDB_Metrics_Advanced.json  MySQL_User_Statistics.json

MongoDB_InMemory.json         MySQL_MyISAM_Metrics.json           PXC_Galera_Cluster_Overview.json

MongoDB_MMAPv1.json           MySQL_Overview.json                 PXC_Galera_Graphs.json

 

拷到宿主機的目錄中：

[root@workstation2017 ~]# docker cp f3e8318c8d7c:/usr/share/percona-dashboards/dashboards /soft/

而後用XSHELL 拷到WINDOW目錄下

在主機WEB界面中

在導入框中加入MySQL_InnoDB_Metrics.json

dashboard ---> import---> MySQL_InnoDB_Metrics.json

便可所有配製完成

 

 

PMM-SERVER 與 PMM-CLIENT 互聯檢測

   gui: http://192.168.79.220:80/prometheus/targets

   command: pmm-admin check-network

            pmm-admin list

EG:

[root@workstation2017 ~]#   pmm-admin check-network

PMM Network Status

Server Address | 192.168.79.220

Client Address | 192.168.79.220

* System Time

NTP Server (0.pool.ntp.org)         | 2017-07-24 22:57:08 -0400 EDT

PMM Server                          | 2017-07-25 02:57:08 +0000 GMT

PMM Client                          | 2017-07-24 22:57:08 -0400 EDT

PMM Server Time Drift               | OK

PMM Client Time Drift               | OK

PMM Client to PMM Server Time Drift | OK

* Connection: Client --> Server

-------------------- -------     

SERVER SERVICE       STATUS      

-------------------- -------     

Consul API           OK

Prometheus API       OK

Query Analytics API  OK

Connection duration | 465.152µs

Request duration    | 1.000382ms

Full round trip     | 1.465534ms

* Connection: Client <-- Server

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

SERVICE TYPE   NAME             REMOTE ENDPOINT       STATUS  HTTPS/TLS  PASSWORD

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

linux:metrics  workstation2017  192.168.79.220:42000  OK      YES        -       

mysql:metrics  workstation2017  192.168.79.220:42002  OK      YES        -