Loki 是一个水平可扩展,高可用性,多租户的日志聚合系统
。它的设计非常经济高效且易于操作,因为它不会为日志内容编制索引,而是为每个日志流编制一组标签。
Loki 组成:
loki : 主服务器,负责存储日志和处理查询
promtail : 代理,负责收集日志并将其发送给 loki
Grafana : Go语言开发的开源数据可视化工具,可以做数据监控和数据统计,带有告警功能
登录官网:loki官网Like Prometheus, but for logs. Contribute to grafana/loki development by creating an account on GitHub.https://github.com/grafana/loki/releases/
----root用户操作
###创建用户
useradd loki
passwd loki
###创建安装目录
mkdir /opt/loki
###授权
chown -R loki:loki /opt/loki
chmod -R 755 /opt/loki
----loki用户操作
###下载二进制包
$ curl -O -L "https://github.com/grafana/loki/releases/download/v2.7.4/loki-linux-amd64.zip"
###解压二进制包
$ unzip "loki-linux-amd64.zip"
### make sure it is executable
$ chmod a+x "loki-linux-amd64"
$ vim loki-local-config.yml
更多配置请参考:https://grafana.com/docs/loki/latest/configuration/
auth_enabled: false
server:
http_listen_port: 3100
grpc_listen_port: 9096
ingester:
lifecycler:
address: 127.0.0.1
ring:
kvstore:
store: inmemory
replication_factor: 1
final_sleep: 0s
chunk_idle_period: 10m
chunk_retain_period: 30s
schema_config:
configs:
- from: 2020-05-15
store: boltdb
object_store: filesystem
schema: v11
index:
prefix: index_
period: 168h
storage_config:
boltdb:
directory: /opt/loki/index
filesystem:
directory: /opt/loki/chunks # 块存储路径
limits_config:
enforce_metric_name: false
reject_old_samples: true # 是否拒绝老样本
reject_old_samples_max_age: 168h # 168小时之前的样本将会被删除
ingestion_rate_mb: 200
ingestion_burst_size_mb: 300
per_stream_rate_limit: 1000MB
max_entries_limit_per_query: 10000
chunk_store_config:
max_look_back_period: 168h # 为避免查询超过保留期的数据,必须小于或等于下方的时间值
table_manager:
retention_deletes_enabled: true # 保留删除开启
retention_period: 168h # 超过168h的块数据将被删除
ruler:
storage:
type: local
local:
directory: /opt/loki/rules
rule_path: /opt/loki/rules-temp
alertmanager_url: http://192.168.0.1:9093 # alertmanager地址
ring:
kvstore:
store: inmemory
enable_api: true
enable_alertmanager_v2: true
$ vim restart-loki.sh
#!/bin/bash
echo "stop loki"
ps -ef | grep loki-linux-amd64 | grep -v grep | awk '{print $2}'| xargs kill -9
echo "Begin start loki"
sleep 1
str=$"\n"
nohup ./loki-linux-amd64 --config.file=loki-local-config.yml &
sstr=$(echo -e $str)
echo $sstr
$chmod +x restart-loki.sh
cd /opt/loki
./restart-loki.sh
[root@localhost loki]# tail -10f nohup.out
level=info ts=2023-03-26T08:40:48.324631438Z caller=checkpoint.go:614 msg="starting checkpoint"
level=info ts=2023-03-26T08:40:48.324805699Z caller=checkpoint.go:339 msg="attempting checkpoint for" dir=wal/checkpoint.000022
level=info ts=2023-03-26T08:45:18.327347566Z caller=checkpoint.go:501 msg="atomic checkpoint finished" old=wal/checkpoint.000022.tmp new=wal/checkpoint.000022
level=info ts=2023-03-26T08:45:18.32797424Z caller=checkpoint.go:572 msg="checkpoint done" time=4m30.002891425s
level=info ts=2023-03-26T08:45:48.325049331Z caller=checkpoint.go:614 msg="starting checkpoint"
level=info ts=2023-03-26T08:45:48.32524561Z caller=checkpoint.go:339 msg="attempting checkpoint for" dir=wal/checkpoint.000023
level=info ts=2023-03-26T08:50:18.327325742Z caller=checkpoint.go:501 msg="atomic checkpoint finished" old=wal/checkpoint.000023.tmp new=wal/checkpoint.000023
level=info ts=2023-03-26T08:50:18.327880793Z caller=checkpoint.go:572 msg="checkpoint done" time=4m30.002463306s
level=info ts=2023-03-26T08:50:48.325234397Z caller=checkpoint.go:614 msg="starting checkpoint"
level=info ts=2023-03-26T08:50:48.325402194Z caller=checkpoint.go:339 msg="attempting checkpoint for" dir=wal/checkpoint.000024
负责收集日志并将其发送给 loki,在需要采集日志的主机上都安装Promtail
mkdir /opt/promtail
chown -R loki:loki /opt/promtail
chmod -R loki:loki /opt/promtail
###下载
curl -O -L "https://github.com/grafana/loki/releases/download/v2.7.4/promtail-linux-amd64.zip"
###解压
unzip loki-linux-amd64.zip # 解压
vim promtail-local-config.yml
server:
http_listen_port: 9080
grpc_listen_port: 0
positions:
filename: /opt/promtail/positions.yaml
clients:
- url: http://10.X.X.125:3100/loki/api/v1/push # 填写好Loki地址
scrape_configs:
- job_name: messagelog
static_configs:
- targets:
- localhost
labels:
#job: messagelog
host: namenode01
__path__: /var/log/messages
- job_name: agentlog
static_configs:
- targets:
- localhost
labels:
#job: agentlog
host: namenode01
__path__: /ssd/ssd0/agentlog/agent.log
- job_name: datanodelog
static_configs:
- targets:
- localhost
labels:
# #job: datanodelog
host: namenode01
__path__: /ssd/ssd0/datanodelog/*datanode*.log
vi restart-promtail.sh
#!/bin/bash
echo "Begin stop promtail"
ps -ef | grep promtail-linux-amd64 | grep -v grep | awk '{print $2}' | xargs kill -9
echo "Begin start promtail...."
nohup ./promtail-linux-amd64 --config.file=promtail-local-config.yml > ./promtail-9080.log 2>&1 &
$chmod +x restart-promtail.sh
[root@localhost promtail]# cd /opt/promtail
[root@localhost promtail]# ./restart-promtail.sh
Begin stop promtail
Begin start promtail....
[root@localhost promtail]#
Grafana : Go语言开发的开源数据可视化工具,可以做数据监控和数据统计,带有告警功能
Download Grafana | Grafana LabsOverview of how to download and install different versions of Grafana on different operating systems.https://grafana.com/grafana/download
###CentOS系统通过rpm方式安装
wget https://dl.grafana.com/enterprise/release/grafana-enterprise-9.4.3-1.x86_64.rpm
sudo yum install grafana-enterprise-9.4.3-1.x86_64.rpm -y
[root@localhost grafana]# rpm -qa|grep grafana
grafana-enterprise-9.4.3-1.x86_64
[root@localhost grafana]#
grafana会占用服务器3000端口
##启动
systemctl start grafana-server.service
##停止
systemctl stop grafana-server.service
##重启
systemctl restart grafana-server.service
##设置开机自启动
systemctl enable grafana-server.service
##状态查看
systemctl status grafana-server.service
Grafana Loki Dashboard for NGINX Service Mesh | Grafana LabsLoki v2+ showcase using JSON NGINX access logs.https://grafana.com/grafana/dashboards/12559-grafana-loki-dashboard-for-nginx-service-mesh/
编辑Grafana配置文件,设置发件人
docker安装的Grafana:vim /home/monitor/grafana/config/grafana.ini
yum安装:vim /etc/grafana/grafana.ini
在[smtp] 标签下 修改配置
# 配置邮件服务器
[smtp]
enabled = true
# 发件服务器
host = smtp.mail.139.com:465
# smtp 密码
password = PAASmail2019
# smtp账号
user = [email protected]
# 发信邮箱
from_address = [email protected]
# 发信人
from_name = Grafana
保存好配置文件之后,重启Grafana
##重启
systemctl restart grafana-server.service
点击页面左边的铃铛,并选择 Contact points
然后选择 Add contact point 创建一个 contact point,填写 Name 和收件人邮箱地址
然后点击Test ,提示成功后,刚填写的邮箱应该就可以收到如下内容的测试邮件
查看邮箱,成功收到邮件
参考:简易日志系统LPG生产环境实践指南 | 坑我已经帮你们踩好了
参考:PLG日志平台搭建: Promtail + Loki + Grafana 全步骤_promtail loki_dxccccccccccc的博客-CSDN博客
参考:Promtail + Loki + Grafana 构建日志监控告警系统(一)_promtail+loki_5D金豆豆的博客-CSDN博客
参考:
PLG日志平台搭建: Promtail + Loki + Grafana 全步骤_promtail loki_dxccccccccccc的博客-CSDN博客
参考:
搭建Loki、Promtail、Grafana轻量级日志系统(centos7)_loki promtail_成钰的博客-CSDN博客
参考:
轻量日志收集系统loki_loki日志收集_wyl9527的博客-CSDN博客
参考:
轻量级日志可视化平台Grafana Loki接入nginx访问日志-阿里云开发者社区
参考:
PLG 云原生日志系统在压测中的落地实践 · TesterHome
参考:
快速部署Grafana日志监控+Nginx封禁IP-阿里云开发者社区
参考:
使用使用grafana中文插件-阿里云开发者社区