TIDB 数据库集群部署

共计 17071 个字符，预计需要花费 43 分钟才能阅读完成。

TiDB 是 PingCAP 公司设计的开源分布式 HTAP (Hybrid Transactional and Analytical Processing) 数据库，结合了传统的 RDBMS 和 NoSQL 的最佳特性。TiDB 兼容 MySQL，支持无限的水平扩展，具备强一致性和高可用性。TiDB 的目标是为 OLTP (Online Transactional Processing) 和 OLAP (Online Analytical Processing) 场景提供一站式的解决方案。

TiDB 具备如下特性：

高度兼容 MySQL

大多数情况下，无需修改代码即可从 MySQL 轻松迁移至 TiDB，分库分表后的 MySQL 集群亦可通过 TiDB 工具进行实时迁移。

水平弹性扩展

通过简单地增加新节点即可实现 TiDB 的水平扩展，按需扩展吞吐或存储，轻松应对高并发、海量数据场景。

分布式事务

TiDB 100% 支持标准的 ACID 事务。

真正金融级高可用

相比于传统主从 (M-S) 复制方案，基于 Raft 的多数派选举协议可以提供金融级的 100% 数据强一致性保证，且在不丢失大多数副本的前提下，可以实现故障的自动恢复 (auto-failover)，无需人工介入。

一站式 HTAP 解决方案

TiDB 作为典型的 OLTP 行存数据库，同时兼具强大的 OLAP 性能，配合 TiSpark，可提供一站式 HTAP 解决方案，一份存储同时处理 OLTP & OLAP，无需传统繁琐的 ETL 过程。

云原生 SQL 数据库

TiDB 是为云而设计的数据库，支持公有云、私有云和混合云，使部署、配置和维护变得十分简单。

TiDB Server

TiDB Server 负责接收 SQL 请求，处理 SQL 相关的逻辑，并通过 PD 找到存储计算所需数据的 TiKV 地址，与 TiKV 交互获取数据，最终返回结果。TiDB Server 是无状态的，其本身并不存储数据，只负责计算，可以无限水平扩展，可以通过负载均衡组件（如 LVS、HAProxy 或 F5）对外提供统一的接入地址。

PD Server

Placement Driver (简称 PD) 是整个集群的管理模块，其主要工作有三个：一是存储集群的元信息（某个 Key 存储在哪个 TiKV 节点）；二是对 TiKV 集群进行调度和负载均衡（如数据的迁移、Raft group leader 的迁移等）；三是分配全局唯一且递增的事务 ID。

PD 是一个集群，需要部署奇数个节点，一般线上推荐至少部署 3 个节点

TiKV Server

TiKV Server 负责存储数据，从外部看 TiKV 是一个分布式的提供事务的 Key-Value 存储引擎。存储数据的基本单位是 Region，每个 Region 负责存储一个 Key Range（从 StartKey 到 EndKey 的左闭右开区间）的数据，每个 TiKV 节点会负责多个 Region。TiKV 使用 Raft 协议做复制，保持数据的一致性和容灾。副本以 Region 为单位进行管理，不同节点上的多个 Region 构成一个 Raft Group，互为副本。数据在多个 TiKV 之间的负载均衡由 PD 调度，这里也是以 Region 为单位进行调度

TiSpark

TiSpark 作为 TiDB 中解决用户复杂 OLAP 需求的主要组件，将 Spark SQL 直接运行在 TiDB 存储层上，同时融合 TiKV 分布式集群的优势，并融入大数据社区生态。至此，TiDB 可以通过一套系统，同时支持 OLTP 与 OLAP，免除用户数据同步的烦恼

创建、查看和删除数据库

CREATE DATABASE db_name [options];
CREATE DATABASE IF NOT EXISTS samp_db;
DROP DATABASE samp_db;
DROP TABLE IF EXISTS person;
CREATE INDEX person_num ON person (number);
ALTER TABLE person ADD INDEX person_num (number);
CREATE UNIQUE INDEX person_num ON person (number);
CREATE USER ‘tiuser’@’localhost’ IDENTIFIED BY ‘123456’;
GRANT SELECT ON samp_db.* TO ‘tiuser’@’localhost’;
SHOW GRANTS for tiuser@localhost;
DROP USER ‘tiuser’@’localhost’;
GRANT ALL PRIVILEGES ON test.* TO ‘xxxx’@’%’ IDENTIFIED BY ‘yyyyy’;
REVOKE ALL PRIVILEGES ON `test`.* FROM ‘genius’@’localhost’;
SHOW GRANTS for ‘root’@’%’;
SELECT Insert_priv FROM mysql.user WHERE user=’test’ AND host=’%’;
FLUSH PRIVILEGES;

使用三台物理机搭建 Tidb 集群，三台机器 ip 为 172.16.5.50，172.16.5.51，172.16.5.10，其中 172.16.5.51 作为中控机。

软件安装如下：

172.16.5.51 TiDB,PD,TiKV

172.16.5.50 TiKV

172.16.5.10 TiKV

安装中控机软件

1 yum -y install epel-release git curl sshpass atop vim htop net-tools
2 yum -y install Python-pip

在中控机上创建 tidb 用户，并生成 ssh key

# 创建 tidb 用户
useradd -m -d /home/tidb tidb && passwd tidb
# 配置 tidb 用户 sudo 权限
visudo
tidb ALL=(ALL) NOPASSWD: ALL
# 使用 tidb 账户生成 ssh key
su tidb && ssh-keygen -t rsa -C mikel@tidb

在中控机器上下载 TiDB-Ansible

1 # 下载 Tidb-Ansible 版本
cd /home/tidb && git clone -b release-2.0 https://github.com/pingcap/tidb-ansible.git
2 # 安装 ansible 及依赖
cd /home/tidb/tidb-ansible/ && pip install -r ./requirements.txt

在中控机上配置部署机器 ssh 互信及 sudo 规则

# 配置 hosts.ini
su tidb && cd /home/tidb/tidb-ansible
vim hosts.ini
[servers]
172.16.5.50
172.16.5.51
172.16.5.52
[all:vars]
username = tidb
ntp_server = pool.ntp.org
# 配置 ssh 互信
ansible-playbook -i hosts.ini create_users.yml -u root -k

在目标机器上安装 ntp 服务

1 # 中控机器上给目标主机安装 ntp 服务
2 cd /home/tidb/tidb-ansible
3 ansible-playbook -i hosts.ini deploy_ntp.yml -u tidb -b

目标机器上调整 cpufreq

1 # 查看 cpupower 调节模式，目前虚拟机不支持，调节 10 服务器 cpupower
2 cpupower frequency-info –governors
3 analyzing CPU 0:
4 available cpufreq governors: Not Available
5 # 配置 cpufreq 调节模式
6 cpupower frequency-set –governor performance

目标机器上添加数据盘 ext4 文件系统挂载

# 创建分区表
parted -s -a optimal /dev/nvme0n1 mklabel gpt — mkpart primary ext4 1 -1
# 手动创建分区
parted dev/sdb
mklabel gpt
mkpart primary 0KB 210GB
# 格式化分区
mkfs.ext4 /dev/sdb
# 查看数据盘分区 UUID
[root@tidb-tikv1 ~]# lsblk -f
NAME FSTYPE LABEL UUID MOUNTPOINT
sda
├─sda1 xfs f41c3b1b-125f-407c-81fa-5197367feb39 /boot
├─sda2 xfs 8119193b-c774-467f-a057-98329c66b3b3 /
├─sda3
└─sda5 xfs 42356bb3-911a-4dc4-b56e-815bafd08db2 /home
sdb ext4 532697e9-970e-49d4-bdba-df386cac34d2
# 分别在三台机器上，编辑 /etc/fstab 文件，添加 nodelalloc 挂载参数
vim /etc/fstab
UUID=8119193b-c774-467f-a057-98329c66b3b3 / xfs defaults 0 0
UUID=f41c3b1b-125f-407c-81fa-5197367feb39 /boot xfs defaults 0 0
UUID=42356bb3-911a-4dc4-b56e-815bafd08db2 /home xfs defaults 0 0
UUID=532697e9-970e-49d4-bdba-df386cac34d2 /data ext4 defaults,nodelalloc,noatime 0 2
# 挂载数据盘
mkdir /data
mount -a
mount -t ext4
/dev/sdb on /data type ext4 (rw,noatime,seclabel,nodelalloc,data=ordered)

分配机器资源，编辑 inventory.ini 文件

# 单机 Tikv 实例
Name HostIP Services
tidb-tikv1 172.16.5.50 PD1, TiDB1, TiKV1
tidb-tikv2 172.16.5.51 PD2, TiKV2
tidb-tikv3 172.16.5.52 PD3, TiKV3
# 编辑 inventory.ini 文件
cd /home/tidb/tidb-ansible
vim inventory.ini
## TiDB Cluster Part
[tidb_servers]
172.16.5.50
172.16.5.51

[tikv_servers]
172.16.5.50
172.16.5.51
172.16.5.52

[pd_servers]
172.16.5.50
172.16.5.51
172.16.5.52

## Monitoring Part
# prometheus and pushgateway servers
[monitoring_servers]
172.16.5.50

# node_exporter and blackbox_exporter servers
[monitored_servers]
172.16.5.50
172.16.5.51
172.16.5.52

[all:vars]
#deploy_dir = /home/tidb/deploy
deploy_dir = /data/deploy
# 检测 ssh 互信
[tidb@tidb-tikv1 tidb-ansible]$ ansible -i inventory.ini all -m shell -a ‘whoami’
172.16.5.51 | SUCCESS | rc=0 >>
tidb
172.16.5.52 | SUCCESS | rc=0 >>
tidb
172.16.5.50 | SUCCESS | rc=0 >>
tidb
# 检测 tidb 用户 sudo 免密码配置
[tidb@tidb-tikv1 tidb-ansible]$ ansible -i inventory.ini all -m shell -a ‘whoami’ -b
172.16.5.52 | SUCCESS | rc=0 >>
root
172.16.5.51 | SUCCESS | rc=0 >>
root
172.16.5.50 | SUCCESS | rc=0 >>
root
# 执行 local_prepare.yml playbook，联网下载 TiDB binary 到中控机
ansible-playbook local_prepare.yml
# 初始化系统环境，修改内核参数
ansible-playbook bootstrap.yml

2.2、安装 Tidb 集群

1 ansible-playbook deploy.yml

2.3、启动 Tidb 集群

1 ansible-playbook start.yml

2.4、测试集群

# 单机 Tikv 实例
Name HostIP Services
tidb-tikv1 172.16.5.50 PD1, TiDB1, TiKV1
tidb-tikv2 172.16.5.51 PD2, TiKV2
tidb-tikv3 172.16.5.52 PD3, TiKV3
# 新增一台 TIDB 节点
添加一个 TiDB 节点（tidb-tikv4），IP 地址为 172.16.5.53
# 编辑 inventory.ini 文件
cd /home/tidb/tidb-ansible
vim inventory.ini
——————start—————————
## TiDB Cluster Part
[tidb_servers]
172.16.5.50
172.16.5.51
172.16.5.53

[tikv_servers]
172.16.5.50
172.16.5.51
172.16.5.52

[pd_servers]
172.16.5.50
172.16.5.51
172.16.5.52

## Monitoring Part
# prometheus and pushgateway servers
[monitoring_servers]
172.16.5.50

# node_exporter and blackbox_exporter servers
[monitored_servers]
172.16.5.50
172.16.5.51
172.16.5.52
172.16.5.53
———————-end——————-
# 拓扑结构如下
Name HostIP Services
tidb-tikv1 172.16.5.50 PD1, TiDB1, TiKV1
tidb-tikv2 172.16.5.51 PD2, TiKV2
tidb-tikv3 172.16.5.52 PD3, TiKV3
tidb-tikv4 172.16.5.53 TiDB2
# 初始化新增节点
ansible-playbook bootstrap.yml -l 172.16.5.53
# 部署新增节点
ansible-playbook deploy.yml -l 172.16.5.53
# 启动新节点服务
ansible-playbook start.yml -l 172.16.5.53
# 更新 Prometheus 配置并重启
ansible-playbook rolling_update_monitor.yml –tags=prometheus

3.2、扩容 PD 节点

# 拓扑结构如下# 单机 Tikv 实例
Name HostIP Services
tidb-tikv1 172.16.5.50 PD1, TiDB1, TiKV1
tidb-tikv2 172.16.5.51 PD2, TiKV2
tidb-tikv3 172.16.5.52 PD3, TiKV3
# 新增一台 PD 节点
添加一个 PD 节点（tidb-pd1），IP 地址为 172.16.5.54
# 编辑 inventory.ini 文件
cd /home/tidb/tidb-ansible
vim inventory.ini
## TiDB Cluster Part
[tidb_servers]
172.16.5.50
172.16.5.51

[tikv_servers]
172.16.5.50
172.16.5.51
172.16.5.52

[pd_servers]
172.16.5.50
172.16.5.51
172.16.5.52
172.16.5.54

## Monitoring Part
# prometheus and pushgateway servers
[monitoring_servers]
172.16.5.50

# node_exporter and blackbox_exporter servers
[monitored_servers]
172.16.5.50
172.16.5.51
172.16.5.52
172.16.5.54
# 拓扑结构如下
Name HostIP Services
tidb-tikv1 172.16.5.50 PD1, TiDB1, TiKV1
tidb-tikv2 172.16.5.51 PD2, TiKV2
tidb-tikv3 172.16.5.52 PD3, TiKV3
tidb-pd1 172.16.5.54 PD4
# 初始化新增节点
ansible-playbook bootstrap.yml -l 172.16.5.54
# 部署新增节点
ansible-playbook deploy.yml -l 172.16.5.54
# 登录新增的 PD 节点，编辑启动脚本：{deploy_dir}/scripts/run_pd.sh
1、移除 –initial-cluster=”xxxx” \ 配置。
2、添加 –join=”http://172.16.10.1:2379″ \，IP 地址（172.16.10.1）可以是集群内现有 PD IP 地址中的任意一个。
3、在新增 PD 节点中手动启动 PD 服务：
{deploy_dir}/scripts/start_pd.sh
4、使用 pd-ctl 检查新节点是否添加成功：
/home/tidb/tidb-ansible/resources/bin/pd-ctl -u “http://172.16.10.1:2379” -d member
# 滚动升级整个集群
ansible-playbook rolling_update.yml
# 更新 Prometheus 配置并重启
ansible-playbook rolling_update_monitor.yml –tags=prometheus

四、tidb 集群测试

4.1、sysbench 基准库测试

sysbench 安装

1 # 二进制安装
2 curl -s https://packagecloud.io/install/repositories/akopytov/sysbench/script.rpm.sh | sudo bash
3 sudo yum -y install sysbench

性能测试

# cpu 性能测试
sysbench –test=cpu –cpu-max-prime=20000 run
———————————-start—————————————-
Number of threads: 1
Initializing random number generator from current time
Prime numbers limit: 20000
Initializing worker threads…
Threads started!
CPU speed:
events per second: 286.71
General statistics:
total time: 10.0004s
total number of events: 2868
Latency (ms):
min: 3.46
avg: 3.49
max: 4.49
95th percentile: 3.55
sum: 9997.23
Threads fairness:
events (avg/stddev): 2868.0000/0.00
execution time (avg/stddev): 9.9972/0.00
———————————–end——————————————-
# 线程测试
sysbench –test=threads –num-threads=64 –thread-yields=100 –thread-locks=2 run
————————————start—————————————–
Number of threads: 64
Initializing random number generator from current time
Initializing worker threads…
Threads started!
General statistics:
total time: 10.0048s
total number of events: 108883
Latency (ms):
min: 0.05
avg: 5.88
max: 49.15
95th percentile: 17.32
sum: 640073.32
Threads fairness:
events (avg/stddev): 1701.2969/36.36
execution time (avg/stddev): 10.0011/0.00
———————————–end—————————————–
# 磁盘 IO 测试
sysbench –test=fileio –num-threads=16 –file-total-size=3G –file-test-mode=rndrw prepare
———————————-start—————————————–
128 files, 24576Kb each, 3072Mb total
Creating files for the test…
Extra file open flags: (none)
Creating file test_file.0
Creating file test_file.1
Creating file test_file.2
Creating file test_file.3
Creating file test_file.4
Creating file test_file.5
Creating file test_file.6
Creating file test_file.7
Creating file test_file.8
Creating file test_file.9
Creating file test_file.10
Creating file test_file.11
Creating file test_file.12
Creating file test_file.13
Creating file test_file.14
Creating file test_file.15
Creating file test_file.16
Creating file test_file.17
Creating file test_file.18
Creating file test_file.19
Creating file test_file.20
Creating file test_file.21
Creating file test_file.22
Creating file test_file.23
Creating file test_file.24
Creating file test_file.25
Creating file test_file.26
Creating file test_file.27
Creating file test_file.28
Creating file test_file.29
Creating file test_file.30
Creating file test_file.31
Creating file test_file.32
Creating file test_file.33
Creating file test_file.34
Creating file test_file.35
Creating file test_file.36
Creating file test_file.37
Creating file test_file.38
Creating file test_file.39
Creating file test_file.40
Creating file test_file.41
Creating file test_file.42
Creating file test_file.43
Creating file test_file.44
Creating file test_file.45
Creating file test_file.46
Creating file test_file.47
Creating file test_file.48
Creating file test_file.49
Creating file test_file.50
Creating file test_file.51
Creating file test_file.52
Creating file test_file.53
Creating file test_file.54
Creating file test_file.55
Creating file test_file.56
Creating file test_file.57
Creating file test_file.58
Creating file test_file.59
Creating file test_file.60
Creating file test_file.61
Creating file test_file.62
Creating file test_file.63
Creating file test_file.64
Creating file test_file.65
Creating file test_file.66
Creating file test_file.67
Creating file test_file.68
Creating file test_file.69
Creating file test_file.70
Creating file test_file.71
Creating file test_file.72
Creating file test_file.73
Creating file test_file.74
Creating file test_file.75
Creating file test_file.76
Creating file test_file.77
Creating file test_file.78
Creating file test_file.79
Creating file test_file.80
Creating file test_file.81
Creating file test_file.82
Creating file test_file.83
Creating file test_file.84
Creating file test_file.85
Creating file test_file.86
Creating file test_file.87
Creating file test_file.88
Creating file test_file.89
Creating file test_file.90
Creating file test_file.91
Creating file test_file.92
Creating file test_file.93
Creating file test_file.94
Creating file test_file.95
Creating file test_file.96
Creating file test_file.97
Creating file test_file.98
Creating file test_file.99
Creating file test_file.100
Creating file test_file.101
Creating file test_file.102
Creating file test_file.103
Creating file test_file.104
Creating file test_file.105
Creating file test_file.106
Creating file test_file.107
Creating file test_file.108
Creating file test_file.109
Creating file test_file.110
Creating file test_file.111
Creating file test_file.112
Creating file test_file.113
Creating file test_file.114
Creating file test_file.115
Creating file test_file.116
Creating file test_file.117
Creating file test_file.118
Creating file test_file.119
Creating file test_file.120
Creating file test_file.121
Creating file test_file.122
Creating file test_file.123
Creating file test_file.124
Creating file test_file.125
Creating file test_file.126
Creating file test_file.127
3221225472 bytes written in 339.76 seconds (9.04 MiB/sec)
———————————-end——————————————
sysbench –test=fileio –num-threads=16 –file-total-size=3G –file-test-mode=rndrw run
———————————-start—————————————–
Number of threads: 16
Initializing random number generator from current time
Extra file open flags: (none)
128 files, 24MiB each
3GiB total file size
Block size 16KiB
Number of IO requests: 0
Read/Write ratio for combined random IO test: 1.50
Periodic FSYNC enabled, calling fsync() each 100 requests.
Calling fsync() at the end of test, Enabled.
Using synchronous I/O mode
Doing random r/w test
Initializing worker threads…
Threads started!
File operations:
reads/s: 299.19
writes/s: 199.46
fsyncs/s: 816.03
Throughput:
read, MiB/s: 4.67
written, MiB/s: 3.12
General statistics:
total time: 10.8270s
total number of events: 12189
Latency (ms):
min: 0.00
avg: 13.14
max: 340.58
95th percentile: 92.42
sum: 160186.15
Threads fairness:
events (avg/stddev): 761.8125/216.01
execution time (avg/stddev): 10.0116/0.01
————————————–end—————————————
sysbench –test=fileio –num-threads=16 –file-total-size=3G –file-test-mode=rndrw cleanup
# 内存测试
sysbench –test=memory –memory-block-size=8k –memory-total-size=4G run
————————————start—————————————–
Number of threads: 1
Initializing random number generator from current time
Running memory speed test with the following options:
block size: 8KiB
total size: 4096MiB
operation: write
scope: global
Initializing worker threads…
Threads started!
Total operations: 524288 (1111310.93 per second)
4096.00 MiB transferred (8682.12 MiB/sec)
General statistics:
total time: 0.4692s
total number of events: 524288
Latency (ms):
min: 0.00
avg: 0.00
max: 0.03
95th percentile: 0.00
sum: 381.39

Threads fairness:
events (avg/stddev): 524288.0000/0.00
execution time (avg/stddev): 0.3814/0.00
————————————-end—————————————

4.2、OLTP 测试

# 登录 tidb 创建测试数据库
mysql -u root -P 4000 -h 172.16.5.50
create database sbtest
# 准备测试数据
sysbench /usr/share/sysbench/oltp_common.lua –mysql-host=172.16.5.50 –mysql-port=4000 –mysql-user=root –tables=20 –table_size=20000000 –threads=100 –max-requests=0 prepare
–tables=20 # 创建 20 个表
–table_size=20000000 # 每个表两千万数据
–threads=100 # 使用 100 个线程数
——————————— 报错信息如下 ——————————————
FATAL: mysql_drv_query() returned error 9001 (PD server timeout[try again later]
2018/11/23 11:23:19.236 log.go:82: [warning] etcdserver: [timed out waiting for read index response]
2018/11/23 14:15:17.329 heartbeat_streams.go:97: [error] [store 1] send keepalive message fail: EOF
2018/11/23 14:14:04.603 leader.go:312: [info] leader is deleted
2018/11/23 14:14:04.603 leader.go:103: [info] pd2 is not etcd leader, skip campaign leader and check later
2018/11/23 14:21:10.071 coordinator.go:570: [info] [region 1093] send schedule command: transfer leader from store 7 to store 2
FATAL: mysql_drv_query() returned error 1105 (Information schema is out of date)
————————————end———————————————–
# 调整线程数为 10，表数量为 10，表数据为 2000000 做测试
sysbench /usr/share/sysbench/oltp_common.lua –mysql-host=172.16.5.50 –mysql-port=4000 –mysql-user=root –tables=1 –table_size=2000000 –threads=10 –max-requests=0 prepare
————————————–start——————————————–
FATAL: mysql_drv_query() returned error 1105 (Information schema is out of date) 超时报错
成功写入 2 张表，其余 8 张表数据并未写满，写好索引
# 对 tidb 集群进行读写测试
sysbench /usr/share/sysbench/oltp_read_write.lua –mysql-host=172.16.5.50 –mysql-port=4000 –mysql-user=root –tables=1 –table_size=2000000 –threads=10 –max-requests=0 run
—————————————-start————————————–
Number of threads: 10
Initializing random number generator from current time
Initializing worker threads…
Threads started!
SQL statistics:
queries performed:
read: 868
write: 62
other: 310
total: 1240
transactions: 62 (5.60 per sec.)
queries: 1240 (112.10 per sec.)
ignored errors: 0 (0.00 per sec.)
reconnects: 0 (0.00 per sec.)
General statistics:
total time: 11.0594s
total number of events: 62
Latency (ms):
min: 944.55
avg: 1757.78
max: 2535.05
95th percentile: 2320.55
sum: 108982.56
Threads fairness:
events (avg/stddev): 6.2000/0.40
execution time (avg/stddev): 10.8983/0.31
————————————end—————————————-
# 使用 mysql 对比测试
mysql -uroot -P 3306 -h 172.15.5.154
create database sbtest
sysbench /usr/share/sysbench/oltp_common.lua –mysql-host=172.16.5.154 –mysql-port=3306 –mysql-user=root –mysql-password=root –tables=20 –table_size=20000000 –threads=10 –max-requests=0 prepare
使用 mysql 做测试未发现报错情况

4.3、业务数据测试

sysbench /usr/share/sysbench/oltp_read_write.lua –mysql-host=172.16.5.50 –mysql-port=4000 –mysql-user=root –tables=20 –table_size=2000000 –threads=10 –max-requests=0 run

：