在知识库中专注阅读，并随时返回相关工具与课程
最近在做 TiDB 的恢复演练，需要在单台 Linux 服务器上部署一套 TiDB 最小的完整拓扑的集群，本文记录一下安装过程。 开始部署 TiDB 集群前，准备一台部署主机，确保其软件满足需求： - 推荐安装 CentOS 7.3 及以上版本 - 运行环境可以支持互联网访问，用于下载 TiDB 及相关软件安装包 注意：TiDB 从 v8.5.1 版本起重新适配 glibc 2.17，恢复了对 C
本章目标：掌握本章核心知识点 前置要求：完成前序章节学习 预计时长：60 分钟
最近在做 TiDB 的恢复演练，需要在单台 Linux 服务器上部署一套 TiDB 最小的完整拓扑的集群，本文记录一下安装过程。
开始部署 TiDB 集群前，准备一台部署主机，确保其软件满足需求：
注意
注意：TiDB 从 v8.5.1 版本起重新适配 glibc 2.17，恢复了对 CentOS Linux 7 的兼容性支持。
最小规模的 TiDB 集群拓扑包含以下实例：
编译和构建 TiDB 所需的依赖库：
下载所需依赖包：
Golang 安装：
1[root@test soft]# tar -C /usr/local -xf go1.25.0.linux-amd64.tar.gz2[root@test ~]# cat<<-\EOF>>/root/.bash_profile3export PATH=$PATH:/usr/local/go/bin4EOF5[root@test ~]# source /root/.bash_profile6[root@test ~]# go version7go version go1.25.0 linux/amd64Rust 安装：
1[root@test soft]# tar -xf rust-1.89.0-x86_64-unknown-linux-gnu.tar.tar2[root@test soft]# cd rust-1.89.0-x86_64-unknown-linux-gnu/3[root@test rust-1.89.0-x86_64-unknown-linux-gnu]# ./install.sh4[root@test ~]# rustc --version5rustc 1.89.0 (29483883e 2025-08-04)sshpass 安装：
1[root@test soft]# tar -xf sshpass-1.10.tar.gz2[root@test soft]# cd sshpass-1.10/3[root@test sshpass-1.10]# ./configure && make && make install4[root@test ~]# sshpass -V5sshpass 1.101[root@test ~]# systemctl stop firewalld.service2[root@test ~]# systemctl disable firewalld.service3[root@test ~]# systemctl status firewalld.service4● firewalld.service - firewalld - dynamic firewall daemon5 Loaded: loaded (/usr/lib/systemd/system/firewalld.service; disabled; vendor preset: enabled)6 Active: inactive (dead)7 Docs: man:firewalld(1)1[root@test ~]# echo "vm.swappiness = 0">> /etc/sysctl.conf2[root@test ~]# swapoff -a3[root@test ~]# sysctl -p4vm.swappiness = 0记得修改 /etc/fstab 配置，注释掉 swap 分区：
1#/dev/mapper/centos-swap swap swap defaults 0 01[root@test ~]# echo never > /sys/kernel/mm/transparent_hugepage/enabled2[root@test ~]# echo never > /sys/kernel/mm/transparent_hugepage/defrag3[root@test ~]# cat<<EOF>>/etc/sysctl.conf4fs.file-max = 10000005net.core.somaxconn = 327686net.ipv4.tcp_tw_recycle = 07net.ipv4.tcp_syncookies = 08vm.overcommit_memory = 19EOF1011[root@test ~]# sysctl -p1213[root@test ~]# cat<<EOF>>/etc/security/limits.conf14tidb soft nofile 100000015tidb hard nofile 100000016tidb soft stack 3276817tidb hard stack 3276818EOF由于模拟多机部署，需要通过 root 用户调大 sshd 服务的连接数限制：
1[root@test ~]# vim /etc/ssh/sshd_config2## 调整 MaxSessions 203[root@test ~]# systemctl restart sshd.service1[root@test ~]# useradd tidb2[root@test ~]# echo "Tidb@123" |passwd tidb --stdin3Changing password for user tidb.4passwd: all authentication tokens updated successfully.5[root@test ~]# cat<<-EOF>>/etc/sudoers6tidb ALL=(ALL) NOPASSWD: ALL7EOF本文是内网环境，不使用官方在线源安装，使用本地镜像源进行部署，本地镜像源部署请参考：TiDB 离线部署 TiUP 组件。
tiup 已部署完成：
1[root@test ~]# tiup mirror show2/root/tidb-community-server-v8.5.3-linux-amd6434[root@test ~]# tiup --version51.16.2 tiup6Go Version: go1.21.137Git Ref: v1.16.28GitHash: 678c52de0c0ef30634b8ba7302a8376caa95d50d创建并启动集群：
1[root@test ~]# cat<<-\EOF>topo.yaml2# # Global variables are applied to all deployments and used as the default value of3# # the deployments if a specific deployment value is missing.4global:5 user: "tidb"6 ssh_port: 111227 deploy_dir: "/data/tidb-deploy"8 data_dir: "/data/tidb-data"910# # Monitored variables are applied to all the machines.11monitored:12 node_exporter_port: 910013 blackbox_exporter_port: 91151415server_configs:16 tidb:17 instance.tidb_slow_log_threshold: 30018 tikv:19 readpool.storage.use-unified-pool: false20 readpool.coprocessor.use-unified-pool: true21 pd:22 replication.enable-placement-rules: true23 replication.location-labels: ["host"]24 tiflash:25 logger.level: "info"2627pd_servers:28 - host: 192.168.31.792930tidb_servers:31 - host: 192.168.31.793233tikv_servers:34 - host: 192.168.31.7935 port: 2016036 status_port: 2018037 config:38 server.labels: { host: "logic-host-1" }3940 - host: 192.168.31.7941 port: 2016142 status_port: 2018143 config:44 server.labels: { host: "logic-host-2" }4546 - host: 192.168.31.7947 port: 2016248 status_port: 2018249 config:50 server.labels: { host: "logic-host-3" }5152tiflash_servers:53 - host: 192.168.31.795455monitoring_servers:56 - host: 192.168.31.795758grafana_servers:59 - host: 192.168.31.7960EOF安装前预检查：
1[root@test ~]# tiup cluster check topo.yaml --user root -p2Input SSH password:34+ Detect CPU Arch Name5 - Detecting node 192.168.31.79 Arch info ... Done67+ Detect CPU OS Name8 - Detecting node 192.168.31.79 OS info ... Done9+ Download necessary tools10 - Downloading check tools for linux/amd64 ... Done11+ Collect basic system information12+ Collect basic system information13 - Getting system info of 192.168.31.79:11122 ... Done14+ Check time zone15 - Checking node 192.168.31.79 ... Done16+ Check system requirements17+ Check system requirements18+ Check system requirements19+ Check system requirements20 - Checking node 192.168.31.79 ... Done21 - Checking node 192.168.31.79 ... Done22 - Checking node 192.168.31.79 ... Done23 - Checking node 192.168.31.79 ... Done24 - Checking node 192.168.31.79 ... Done25 - Checking node 192.168.31.79 ... Done26 - Checking node 192.168.31.79 ... Done27 - Checking node 192.168.31.79 ... Done28 - Checking node 192.168.31.79 ... Done29+ Cleanup check files30 - Cleanup check files on 192.168.31.79:11122 ... Done31Node Check Result Message32---- ----- ------ -------33192.168.31.79 os-version Fail CentOS Linux 7 (Core) 7.9.2009 not supported, use version 9 or higher34192.168.31.79 cpu-cores Pass number of CPU cores / threads: 435192.168.31.79 ntp Warn The NTPd daemon may be not start36192.168.31.79 disk Warn mount point /data does not have 'noatime' option set37192.168.31.79 selinux Pass SELinux is disabled38192.168.31.79 thp Pass THP is disabled39192.168.31.79 command Pass numactl: policy: default40192.168.31.79 cpu-governor Warn Unable to determine current CPU frequency governor policy41192.168.31.79 memory Pass memory size is 8192MB42192.168.31.79 network Pass network speed of ens192 is 10000MB43192.168.31.79 disk Fail multiple components tikv:/data/tidb-data/tikv-20160,tikv:/data/tidb-data/tikv-20161,tikv:/data/tidb-data/tikv-20162,tiflash:/data/tidb-data/tiflash-9000 are using the same partition 192.168.31.79:/data as data dir44192.168.31.79 disk Fail mount point /data does not have 'nodelalloc' option set部署集群：
1[root@test ~]# tiup cluster deploy lucifer v8.5.3 topo.yaml --user root -p2Input SSH password:34+ Detect CPU Arch Name5 - Detecting node 192.168.31.79 Arch info ... Done67+ Detect CPU OS Name8 - Detecting node 192.168.31.79 OS info ... Done9Please confirm your topology:10Cluster type: tidb11Cluster name: lucifer12Cluster version: v8.5.313Role Host Ports OS/Arch Directories14---- ---- ----- ------- -----------15pd 192.168.31.79 2379/2380 linux/x86_64 /data/tidb-deploy/pd-2379,/data/tidb-data/pd-237916tikv 192.168.31.79 20160/20180 linux/x86_64 /data/tidb-deploy/tikv-20160,/data/tidb-data/tikv-2016017tikv 192.168.31.79 20161/20181 linux/x86_64 /data/tidb-deploy/tikv-20161,/data/tidb-data/tikv-2016118tikv 192.168.31.79 20162/20182 linux/x86_64 /data/tidb-deploy/tikv-20162,/data/tidb-data/tikv-2016219tidb 192.168.31.79 4000/10080 linux/x86_64 /data/tidb-deploy/tidb-400020tiflash 192.168.31.79 9000/3930/20170/20292/8234/8123 linux/x86_64 /data/tidb-deploy/tiflash-9000,/data/tidb-data/tiflash-900021prometheus 192.168.31.79 9090/12020 linux/x86_64 /data/tidb-deploy/prometheus-9090,/data/tidb-data/prometheus-909022grafana 192.168.31.79 3000 linux/x86_64 /data/tidb-deploy/grafana-300023Attention:24 1. If the topology is not what you expected, check your yaml file.25 2. Please confirm there is no port/directory conflicts in same host.26Do you want to continue? [y/N]: (default=N) y27+ Generate SSH keys ... Done28+ Download TiDB components29 - Download pd:v8.5.3 (linux/amd64) ... Done30 - Download tikv:v8.5.3 (linux/amd64) ... Done31 - Download tidb:v8.5.3 (linux/amd64) ... Done32 - Download tiflash:v8.5.3 (linux/amd64) ... Done33 - Download prometheus:v8.5.3 (linux/amd64) ... Done34 - Download grafana:v8.5.3 (linux/amd64) ... Done35 - Download node_exporter: (linux/amd64) ... Done36 - Download blackbox_exporter: (linux/amd64) ... Done37+ Initialize target host environments38 - Prepare 192.168.31.79:11122 ... Done39+ Deploy TiDB instance40 - Copy pd -> 192.168.31.79 ... Done41 - Copy tikv -> 192.168.31.79 ... Done42 - Copy tikv -> 192.168.31.79 ... Done43 - Copy tikv -> 192.168.31.79 ... Done44 - Copy tidb -> 192.168.31.79 ... Done45 - Copy tiflash -> 192.168.31.79 ... Done46 - Copy prometheus -> 192.168.31.79 ... Done47 - Copy grafana -> 192.168.31.79 ... Done48 - Deploy node_exporter -> 192.168.31.79 ... Done49 - Deploy blackbox_exporter -> 192.168.31.79 ... Done50+ Copy certificate to remote host51+ Init instance configs52 - Generate config pd -> 192.168.31.79:2379 ... Done53 - Generate config tikv -> 192.168.31.79:20160 ... Done54 - Generate config tikv -> 192.168.31.79:20161 ... Done55 - Generate config tikv -> 192.168.31.79:20162 ... Done56 - Generate config tidb -> 192.168.31.79:4000 ... Done57 - Generate config tiflash -> 192.168.31.79:9000 ... Done58 - Generate config prometheus -> 192.168.31.79:9090 ... Done59 - Generate config grafana -> 192.168.31.79:3000 ... Done60+ Init monitor configs61 - Generate config node_exporter -> 192.168.31.79 ... Done62 - Generate config blackbox_exporter -> 192.168.31.79 ... Done63Enabling component pd64 Enabling instance 192.168.31.79:237965 Enable instance 192.168.31.79:2379 success66Enabling component tikv67 Enabling instance 192.168.31.79:2016268 Enabling instance 192.168.31.79:2016069 Enabling instance 192.168.31.79:2016170 Enable instance 192.168.31.79:20162 success71 Enable instance 192.168.31.79:20161 success72 Enable instance 192.168.31.79:20160 success73Enabling component tidb74 Enabling instance 192.168.31.79:400075 Enable instance 192.168.31.79:4000 success76Enabling component tiflash77 Enabling instance 192.168.31.79:900078 Enable instance 192.168.31.79:9000 success79Enabling component prometheus80 Enabling instance 192.168.31.79:909081 Enable instance 192.168.31.79:9090 success82Enabling component grafana83 Enabling instance 192.168.31.79:300084 Enable instance 192.168.31.79:3000 success85Enabling component node_exporter86 Enabling instance 192.168.31.7987 Enable 192.168.31.79 success88Enabling component blackbox_exporter89 Enabling instance 192.168.31.7990 Enable 192.168.31.79 success91Cluster `lucifer` deployed successfully, you can start it with command: `tiup cluster start lucifer --init`启动集群：
1[root@test ~]# tiup cluster start lucifer --init2Starting cluster lucifer...3+ [ Serial ] - SSHKeySet: privateKey=/root/.tiup/storage/cluster/clusters/lucifer/ssh/id_rsa, publicKey=/root/.tiup/storage/cluster/clusters/lucifer/ssh/id_rsa.pub4+ [Parallel] - UserSSH: user=tidb, host=192.168.31.795+ [Parallel] - UserSSH: user=tidb, host=192.168.31.796+ [Parallel] - UserSSH: user=tidb, host=192.168.31.797+ [Parallel] - UserSSH: user=tidb, host=192.168.31.798+ [Parallel] - UserSSH: user=tidb, host=192.168.31.799+ [Parallel] - UserSSH: user=tidb, host=192.168.31.7910+ [Parallel] - UserSSH: user=tidb, host=192.168.31.7911+ [Parallel] - UserSSH: user=tidb, host=192.168.31.7912+ [ Serial ] - StartCluster13Starting component pd14 Starting instance 192.168.31.79:237915 Start instance 192.168.31.79:2379 success16Starting component tikv17 Starting instance 192.168.31.79:2016218 Starting instance 192.168.31.79:2016019 Starting instance 192.168.31.79:2016120 Start instance 192.168.31.79:20162 success21 Start instance 192.168.31.79:20161 success22 Start instance 192.168.31.79:20160 success23Starting component tidb24 Starting instance 192.168.31.79:400025 Start instance 192.168.31.79:4000 success26Starting component tiflash27 Starting instance 192.168.31.79:900028 Start instance 192.168.31.79:9000 success29Starting component prometheus30 Starting instance 192.168.31.79:909031 Start instance 192.168.31.79:9090 success32Starting component grafana33 Starting instance 192.168.31.79:300034 Start instance 192.168.31.79:3000 success35Starting component node_exporter36 Starting instance 192.168.31.7937 Start 192.168.31.79 success38Starting component blackbox_exporter39 Starting instance 192.168.31.7940 Start 192.168.31.79 success41+ [ Serial ] - UpdateTopology: cluster=lucifer42Started cluster `lucifer` successfully43The root password of TiDB database has been changed.44The new password is: 'm+92G0Q3eNR4^6cq*@'.45Copy and record it to somewhere safe, it is only displayed once, and will not be stored.46The generated password can NOT be get and shown again.查看集群：
1[root@test ~]# tiup cluster list2Name User Version Path PrivateKey3---- ---- ------- ---- ----------4lucifer tidb v8.5.3 /root/.tiup/storage/cluster/clusters/lucifer /root/.tiup/storage/cluster/clusters/lucifer/ssh/id_rsa检查集群状态：
1[root@test ~]# tiup cluster display lucifer2Cluster type: tidb3Cluster name: lucifer4Cluster version: v8.5.35Deploy user: tidb6SSH type: builtin7Dashboard URL: http://192.168.31.79:2379/dashboard8Dashboard URLs: http://192.168.31.79:2379/dashboard9Grafana URL: http://192.168.31.79:300010ID Role Host Ports OS/Arch Status Data Dir Deploy Dir11-- ---- ---- ----- ------- ------ -------- ----------12192.168.31.79:3000 grafana 192.168.31.79 3000 linux/x86_64 Up - /data/tidb-deploy/grafana-300013192.168.31.79:2379 pd 192.168.31.79 2379/2380 linux/x86_64 Up|L|UI /data/tidb-data/pd-2379 /data/tidb-deploy/pd-237914192.168.31.79:9090 prometheus 192.168.31.79 9090/12020 linux/x86_64 Up /data/tidb-data/prometheus-9090 /data/tidb-deploy/prometheus-909015192.168.31.79:4000 tidb 192.168.31.79 4000/10080 linux/x86_64 Up - /data/tidb-deploy/tidb-400016192.168.31.79:9000 tiflash 192.168.31.79 9000/3930/20170/20292/8234/8123 linux/x86_64 Up /data/tidb-data/tiflash-9000 /data/tidb-deploy/tiflash-900017192.168.31.79:20160 tikv 192.168.31.79 20160/20180 linux/x86_64 Up /data/tidb-data/tikv-20160 /data/tidb-deploy/tikv-2016018192.168.31.79:20161 tikv 192.168.31.79 20161/20181 linux/x86_64 Up /data/tidb-data/tikv-20161 /data/tidb-deploy/tikv-2016119192.168.31.79:20162 tikv 192.168.31.79 20162/20182 linux/x86_64 Up /data/tidb-data/tikv-20162 /data/tidb-deploy/tikv-2016220Total nodes: 8TiDB 兼容 MySQL 协议，故需要 MySQL 客户端连接，则需安装 MySQL 客户端，Linux7 版本的系统默认自带安装了 MariaDB，需要先清理：
1[root@test ~]# rpm -e --nodeps $(rpm -qa | grep mariadb)找个有网的环境下载：
1[root@lucifer ~]# wget https://repo.mysql.com/RPM-GPG-KEY-mysql-20232[root@lucifer ~]# wget http://dev.mysql.com/get/mysql80-community-release-el7-10.noarch.rpm安装 MySQL 客户端：
1[root@test ~]# yum -y install mysql80-community-release-el7-10.noarch.rpm2[root@test ~]# rpm --import RPM-GPG-KEY-mysql-20233[root@test ~]# yum -y install mysql连接数据库：
1## 这里的 root 初始密码在 tidb 集群初始化时日志中输出的密码 m+92G0Q3eNR4^6cq*@2[root@test ~]# mysql -h 192.168.31.79 -P 4000 -uroot –p3mysql> show databases;修改初始 root 密码：
1mysql> use mysql2mysql> alter user 'root'@'%' identified by 'tidb';集群监控：
至此，TiDB 单机集群部署完成，可用于开发测试和学习研究。生产环境建议参考官方推荐的多机部署方案。