安装部署
拉取镜像
docker pull clickhouse/clickhouse-server启动实例
创建数据目录并启动
bash
#!/bin/bash
docker run --restart=always -d -P\
-e CLICKHOUSE_DEFAULT_ACCESS_MANAGEMENT=1 \
-e CLICKHOUSE_USER=default \
-e CLICKHOUSE_PASSWORD=nQczEy2EWJ7w \
-v "/data/clickhouse/data:/var/lib/clickhouse/" \
-v "/data/clickhouse/logs:/var/log/clickhouse-server/" \
--name clickhouse-server --ulimit nofile=262144:262144 clickhouse/clickhouse-serverYou may also want to mount:
/etc/clickhouse-server/config.d/*.xml- files with server configuration adjustments/etc/clickhouse-server/users.d/*.xml- files with user settings adjustments/docker-entrypoint-initdb.d/- folder with database initialization scripts (see below).
查看自动分配的端口
[root@localhost ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
0bab25db35bf clickhouse/clickhouse-server "/entrypoint.sh" 1 second ago Up 1 second 0.0.0.0:32773->8123/tcp, :::32773->8123/tcp, 0.0.0.0:32772->9000/tcp, :::32772->9000/tcp, 0.0.0.0:32771->9009/tcp, :::32771->9009/tcp clickhouse-server端口说明
The container exposes port 8123 for the HTTP interface and port 9000 for the native client.
初步使用
容器内CLI客户端连接
SQL语法与MySQL基本相同
[root@localhost ~]# docker exec -it clickhouse-server clickhouse-client
ClickHouse client version 25.8.4.13 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 25.8.4.
Warnings:
* Linux transparent hugepages are set to "always". Check /sys/kernel/mm/transparent_hugepage/enabled
0bab25db35bf :) SELECT 'Hello, ClickHouse!'
SELECT 'Hello, ClickHouse!'
Query id: 78e7cfaa-5118-44d9-ae45-9104eadf1149
┌─'Hello, ClickHouse!'─┐
1. │ Hello, ClickHouse! │
└──────────────────────┘
1 row in set. Elapsed: 0.001 sec.
0bab25db35bf :) exit
Bye.
[root@localhost ~]#说明： ClickHouse 服务器配置了密码认证，在 ClickHouse 中，默认配置下，本地连接不需密码。
容器外使用curl连接
[root@localhost ~]# echo "SELECT 'Hello, ClickHouse!'" | curl 'http://default:nQczEy2EWJ7w@localhost:32773/?query=' -s --data-binary @-
echo "SELECT 'Hello, ClickHouse'" | curl 'http://default:nQczEy2EWJ7w@localhost:32773/?query=' -s --data-binary @-
Hello, ClickHouse修改密码方法
# 进入 ClickHouse 客户端
docker exec -it clickhouse-server clickhouse-client
# 执行修改密码命令
ALTER USER default IDENTIFIED BY 'new_secure_password';模拟数据分析场景
我们来模拟写入一款游戏Command Modern Operations的一场推演中产生的所有事件信息，我们使用uuid来标记一个推演场次，即每一场推演的所有事件数据，uuid一样。根据事件产生大量事件日志，比如红方f22战斗机升空，红方发射xx型号导弹，xx型号导弹命中蓝方xx目标造成xx点损伤。模拟数据尽量考虑cmo游戏中的各种武器装备。请提供python程序，随机生成1百万个随机事件，每5000个事件一组，批量写入到clickhouse。
创建表
CREATE TABLE cmo_events
(
event_id UUID,
session_uuid UUID,
event_time DateTime,
event_type String,
side Enum8('RED' = 1, 'BLUE' = 2),
unit_type String,
unit_name String,
weapon_system Nullable(String),
target_type Nullable(String),
target_name Nullable(String),
damage_points Nullable(Int32),
location_lat Float64,
location_lon Float64,
description String
) ENGINE = MergeTree()
ORDER BY (session_uuid, event_time);参考：
[root@localhost ~]# docker exec -it clickhouse-server clickhouse-client
ClickHouse client version 25.8.4.13 (official build).
Connecting to localhost:9000 as user default.
Connected to ClickHouse server version 25.8.4.
Warnings:
* Linux transparent hugepages are set to "always". Check /sys/kernel/mm/transparent_hugepage/enabled
0bab25db35bf :) CREATE TABLE cmo_events
(
event_id UUID,
session_uuid UUID,
event_time DateTime,
event_type String,
side Enum8('RED' = 1, 'BLUE' = 2),
unit_type String,
unit_name String,
weapon_system Nullable(String),
target_type Nullable(String),
target_name Nullable(String),
damage_points Nullable(Int32),
location_lat Float64,
location_lon Float64,
description String
) ENGINE = MergeTree()
ORDER BY (session_uuid, event_time);
CREATE TABLE cmo_events
(
`event_id` UUID,
`session_uuid` UUID,
`event_time` DateTime,
`event_type` String,
`side` Enum8('RED' = 1, 'BLUE' = 2),
`unit_type` String,
`unit_name` String,
`weapon_system` Nullable(String),
`target_type` Nullable(String),
`target_name` Nullable(String),
`damage_points` Nullable(Int32),
`location_lat` Float64,
`location_lon` Float64,
`description` String
)
ENGINE = MergeTree
ORDER BY (session_uuid, event_time)
Query id: 3a225705-8067-4803-adfe-1f5916b1cc5f
Ok.
0 rows in set. Elapsed: 0.006 sec.
0bab25db35bf :) show tables;
SHOW TABLES
Query id: 8b2ab058-53ab-4ea4-bc1b-a357be3ee4a2
┌─name───────┐
1. │ cmo_events │
└────────────┘
1 row in set. Elapsed: 0.002 sec.
0bab25db35bf :)python数据生成及写入程序
安装驱动
pip install clickhouse-drivergenerate_cmo_events.py
import uuid
import random
import time
from datetime import datetime, timedelta
from clickhouse_driver import Client
from clickhouse_driver.errors import Error
import logging
# 设置日志
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
# ClickHouse连接配置
CH_HOST = 'localhost'
CH_PORT = 32772 # 默认原生TCP端口
CH_USER = 'default'
CH_PASSWORD = 'nQczEy2EWJ7w' # 如果设置了密码，请填写
CH_DATABASE = 'default'
CH_TABLE = 'cmo_events'
# 模拟数据配置
RED_UNITS = [
"F-22A Raptor", "Su-57 Felon", "J-20 Mighty Dragon", "T-14 Armata",
"S-400 Triumf", "Borei-class SSBN", "Yasen-class SSN", "Kirov-class Battlecruiser"
]
BLUE_UNITS = [
"F-35 Lightning II", "F/A-18E/F Super Hornet", "Arleigh Burke-class DDG",
"Virginia-class SSN", "Zumwalt-class DDG", "M1A2 Abrams", "Patriot SAM",
"Trident II D5 SLBM"
]
WEAPON_SYSTEMS = [
"AIM-120D AMRAAM", "R-77 Adder", "PL-15", "AIM-9X Sidewinder",
"R-73 Archer", "Tomahawk Cruise Missile", "Kalibr Cruise Missile",
"DF-21D ASBM", "Standard Missile-6", "Torpedo MK-48", "Torpedo 53-65"
]
EVENT_TYPES = [
"TAKEOFF", "LANDING", "WEAPON_LAUNCH", "WEAPON_HIT", "WEAPON_MISS",
"DETECTION", "FUEL_STATUS", "DAMAGE_ASSESSMENT", "MISSION_UPDATE",
"COMMUNICATION", "SENSOR_ACTIVATION", "COUNTERMEASURE_DEPLOYED"
]
TARGET_TYPES = [
"AIRBASE", "AIRCRAFT", "WARSHIP", "SUBMAIRNE", "GROUND_VEHICLE",
"SAM_SITE", "RADAR_SITE", "COMMAND_CENTER", "INFRASTRUCTURE"
]
# 创建ClickHouse客户端
def create_clickhouse_client():
try:
client = Client(
host=CH_HOST,
port=CH_PORT,
user=CH_USER,
password=CH_PASSWORD,
database=CH_DATABASE
)
logger.info("Successfully connected to ClickHouse")
return client
except Error as e:
logger.error(f"Failed to connect to ClickHouse: {e}")
raise
# 创建表（如果不存在）
def create_table_if_not_exists(client):
create_table_query = f"""
CREATE TABLE IF NOT EXISTS {CH_TABLE}
(
event_id UUID,
session_uuid UUID,
event_time DateTime,
event_type String,
side Enum8('RED' = 1, 'BLUE' = 2),
unit_type String,
unit_name String,
weapon_system Nullable(String),
target_type Nullable(String),
target_name Nullable(String),
damage_points Nullable(Int32),
location_lat Float64,
location_lon Float64,
description String
) ENGINE = MergeTree()
ORDER BY (session_uuid, event_time)
"""
try:
client.execute(create_table_query)
logger.info(f"Table {CH_TABLE} created or already exists")
except Error as e:
logger.error(f"Failed to create table: {e}")
raise
# 生成随机事件
def generate_event(session_uuid, start_time):
event_time = start_time + timedelta(seconds=random.randint(0, 3600*6)) # 6小时内的随机时间
side = random.choice(["RED", "BLUE"])
if side == "RED":
unit_type = random.choice(RED_UNITS)
unit_name = f"Red {unit_type} #{random.randint(1, 12)}"
else:
unit_type = random.choice(BLUE_UNITS)
unit_name = f"Blue {unit_type} #{random.randint(1, 12)}"
event_type = random.choice(EVENT_TYPES)
# 根据事件类型设置特定字段
weapon_system = random.choice(WEAPON_SYSTEMS) if event_type in ["WEAPON_LAUNCH", "WEAPON_HIT", "WEAPON_MISS"] else None
if event_type in ["WEAPON_HIT", "WEAPON_MISS", "DAMAGE_ASSESSMENT"]:
target_side = "BLUE" if side == "RED" else "RED"
target_type = random.choice(TARGET_TYPES)
target_name = f"{target_side} {target_type} #{random.randint(1, 20)}"
damage_points = random.randint(10, 100) if event_type in ["WEAPON_HIT", "DAMAGE_ASSESSMENT"] else None
else:
target_type = None
target_name = None
damage_points = None
# 生成位置数据 (模拟在某个区域内的随机位置)
location_lat = 30.0 + random.uniform(-5, 5)
location_lon = 120.0 + random.uniform(-5, 5)
# 生成描述
if event_type == "TAKEOFF":
description = f"{unit_name} took off from base"
elif event_type == "LANDING":
description = f"{unit_name} landed at base"
elif event_type == "WEAPON_LAUNCH":
description = f"{unit_name} launched {weapon_system}"
elif event_type == "WEAPON_HIT":
description = f"{weapon_system} from {unit_name} hit {target_name} causing {damage_points} damage"
elif event_type == "WEAPON_MISS":
description = f"{weapon_system} from {unit_name} missed {target_name}"
elif event_type == "DETECTION":
description = f"{unit_name} detected unknown contact"
else:
description = f"{unit_name} {event_type.lower().replace('_', ' ')}"
return [
uuid.uuid4(), # event_id
session_uuid, # session_uuid
event_time, # event_time
event_type, # event_type
side, # side
unit_type, # unit_type
unit_name, # unit_name
weapon_system, # weapon_system
target_type, # target_type
target_name, # target_name
damage_points, # damage_points
location_lat, # location_lat
location_lon, # location_lon
description # description
]
# 批量生成并插入数据
def generate_and_insert_data(client, num_events=1000000, batch_size=5000):
session_uuid = uuid.uuid4()
start_time = datetime.now() - timedelta(hours=6) # 从6小时前开始
logger.info(f"Generating {num_events} events for session {session_uuid}")
total_start_time = time.time()
generated_events = 0
# 准备插入查询
insert_query = f"""
INSERT INTO {CH_TABLE} (
event_id, session_uuid, event_time, event_type, side,
unit_type, unit_name, weapon_system, target_type, target_name,
damage_points, location_lat, location_lon, description
) VALUES
"""
# 分批次生成和插入数据
for batch_num in range(0, num_events, batch_size):
batch_start_time = time.time()
batch_data = []
current_batch_size = min(batch_size, num_events - batch_num)
# 生成批次数据
for i in range(current_batch_size):
event = generate_event(session_uuid, start_time)
batch_data.append(event)
# 插入数据到ClickHouse
try:
client.execute(insert_query, batch_data)
generated_events += current_batch_size
batch_time = time.time() - batch_start_time
logger.info(f"Inserted batch {batch_num//batch_size + 1} with {current_batch_size} events "
f"in {batch_time:.2f} seconds "
f"({current_batch_size/batch_time:.2f} events/sec)")
except Error as e:
logger.error(f"Failed to insert batch {batch_num//batch_size + 1}: {e}")
# 可以选择重试或继续
total_time = time.time() - total_start_time
logger.info(f"Total: Inserted {generated_events} events in {total_time:.2f} seconds "
f"({generated_events/total_time:.2f} events/sec)")
return generated_events, total_time
if __name__ == "__main__":
# 安装依赖: pip install clickhouse-driver
try:
# 创建ClickHouse客户端
client = create_clickhouse_client()
# 创建表（如果不存在）
create_table_if_not_exists(client)
# 生成并插入数据
events_count, total_time = generate_and_insert_data(client, 1000000, 5000)
logger.info(f"Successfully inserted {events_count} events in {total_time:.2f} seconds")
except Exception as e:
logger.error(f"Program failed with error: {e}")运行日志（性能优秀）
通过日志，可以看到写入性能非常炸裂！
[root@localhost clickhouse-sample]# python3 generate_cmo_events.py
INFO:__main__:Successfully connected to ClickHouse
INFO:__main__:Table cmo_events created or already exists
INFO:__main__:Generating 1000000 events for session 09cfee87-3b10-41cb-a995-9bb351944997
INFO:__main__:Inserted batch 1 with 5000 events in 0.10 seconds (52522.95 events/sec)
INFO:__main__:Inserted batch 2 with 5000 events in 0.08 seconds (60039.74 events/sec)
INFO:__main__:Inserted batch 3 with 5000 events in 0.09 seconds (56199.96 events/sec)
INFO:__main__:Inserted batch 4 with 5000 events in 0.08 seconds (59493.17 events/sec)
INFO:__main__:Inserted batch 5 with 5000 events in 0.08 seconds (58916.54 events/sec)
INFO:__main__:Inserted batch 6 with 5000 events in 0.09 seconds (58321.62 events/sec)
INFO:__main__:Inserted batch 7 with 5000 events in 0.09 seconds (58659.95 events/sec)
INFO:__main__:Inserted batch 8 with 5000 events in 0.08 seconds (61833.34 events/sec)
INFO:__main__:Inserted batch 9 with 5000 events in 0.08 seconds (61569.03 events/sec)
INFO:__main__:Inserted batch 10 with 5000 events in 0.08 seconds (60840.98 events/sec)
INFO:__main__:Inserted batch 11 with 5000 events in 0.08 seconds (61788.53 events/sec)
INFO:__main__:Inserted batch 12 with 5000 events in 0.08 seconds (61649.57 events/sec)
INFO:__main__:Inserted batch 13 with 5000 events in 0.10 seconds (49049.30 events/sec)
INFO:__main__:Inserted batch 14 with 5000 events in 0.08 seconds (62063.36 events/sec)
INFO:__main__:Inserted batch 15 with 5000 events in 0.08 seconds (61636.71 events/sec)
INFO:__main__:Inserted batch 16 with 5000 events in 0.08 seconds (59425.57 events/sec)
INFO:__main__:Inserted batch 17 with 5000 events in 0.08 seconds (61817.67 events/sec)
INFO:__main__:Inserted batch 18 with 5000 events in 0.08 seconds (61878.77 events/sec)
INFO:__main__:Inserted batch 19 with 5000 events in 0.08 seconds (61121.50 events/sec)
INFO:__main__:Inserted batch 20 with 5000 events in 0.08 seconds (62007.94 events/sec)
INFO:__main__:Inserted batch 21 with 5000 events in 0.08 seconds (61800.18 events/sec)
INFO:__main__:Inserted batch 22 with 5000 events in 0.08 seconds (61397.97 events/sec)
INFO:__main__:Inserted batch 23 with 5000 events in 0.08 seconds (61920.43 events/sec)
INFO:__main__:Inserted batch 24 with 5000 events in 0.08 seconds (61823.68 events/sec)
INFO:__main__:Inserted batch 25 with 5000 events in 0.08 seconds (61918.23 events/sec)
INFO:__main__:Inserted batch 26 with 5000 events in 0.08 seconds (59881.16 events/sec)
INFO:__main__:Inserted batch 27 with 5000 events in 0.08 seconds (61966.35 events/sec)
INFO:__main__:Inserted batch 28 with 5000 events in 0.08 seconds (62225.78 events/sec)
INFO:__main__:Inserted batch 29 with 5000 events in 0.08 seconds (61433.76 events/sec)
INFO:__main__:Inserted batch 30 with 5000 events in 0.08 seconds (62425.83 events/sec)
INFO:__main__:Inserted batch 31 with 5000 events in 0.08 seconds (61381.26 events/sec)
INFO:__main__:Inserted batch 32 with 5000 events in 0.10 seconds (49932.31 events/sec)
INFO:__main__:Inserted batch 33 with 5000 events in 0.09 seconds (57886.16 events/sec)
INFO:__main__:Inserted batch 34 with 5000 events in 0.08 seconds (59527.11 events/sec)
INFO:__main__:Inserted batch 35 with 5000 events in 0.09 seconds (58397.13 events/sec)
INFO:__main__:Inserted batch 36 with 5000 events in 0.08 seconds (59672.44 events/sec)
INFO:__main__:Inserted batch 37 with 5000 events in 0.08 seconds (59633.92 events/sec)
INFO:__main__:Inserted batch 38 with 5000 events in 0.08 seconds (60850.33 events/sec)
INFO:__main__:Inserted batch 39 with 5000 events in 0.08 seconds (60351.09 events/sec)
INFO:__main__:Inserted batch 40 with 5000 events in 0.08 seconds (60006.58 events/sec)
INFO:__main__:Inserted batch 41 with 5000 events in 0.09 seconds (58819.38 events/sec)
INFO:__main__:Inserted batch 42 with 5000 events in 0.14 seconds (35813.80 events/sec)
INFO:__main__:Inserted batch 43 with 5000 events in 0.12 seconds (42547.30 events/sec)
INFO:__main__:Inserted batch 44 with 5000 events in 0.09 seconds (57668.41 events/sec)
INFO:__main__:Inserted batch 45 with 5000 events in 0.09 seconds (58709.22 events/sec)
INFO:__main__:Inserted batch 46 with 5000 events in 0.08 seconds (60861.28 events/sec)
INFO:__main__:Inserted batch 47 with 5000 events in 0.08 seconds (61517.56 events/sec)
INFO:__main__:Inserted batch 48 with 5000 events in 0.08 seconds (60536.10 events/sec)
INFO:__main__:Inserted batch 49 with 5000 events in 0.08 seconds (61419.19 events/sec)
INFO:__main__:Inserted batch 50 with 5000 events in 0.08 seconds (60751.97 events/sec)
INFO:__main__:Inserted batch 51 with 5000 events in 0.08 seconds (61520.45 events/sec)
INFO:__main__:Inserted batch 52 with 5000 events in 0.08 seconds (60766.41 events/sec)
INFO:__main__:Inserted batch 53 with 5000 events in 0.10 seconds (49802.47 events/sec)
INFO:__main__:Inserted batch 54 with 5000 events in 0.09 seconds (55050.87 events/sec)
INFO:__main__:Inserted batch 55 with 5000 events in 0.08 seconds (60076.37 events/sec)
INFO:__main__:Inserted batch 56 with 5000 events in 0.08 seconds (60173.77 events/sec)
INFO:__main__:Inserted batch 57 with 5000 events in 0.08 seconds (60918.39 events/sec)
INFO:__main__:Inserted batch 58 with 5000 events in 0.08 seconds (60296.26 events/sec)
INFO:__main__:Inserted batch 59 with 5000 events in 0.08 seconds (61255.75 events/sec)
INFO:__main__:Inserted batch 60 with 5000 events in 0.08 seconds (59838.10 events/sec)
INFO:__main__:Inserted batch 61 with 5000 events in 0.08 seconds (60690.09 events/sec)
INFO:__main__:Inserted batch 62 with 5000 events in 0.08 seconds (59796.30 events/sec)
INFO:__main__:Inserted batch 63 with 5000 events in 0.08 seconds (61230.36 events/sec)
INFO:__main__:Inserted batch 64 with 5000 events in 0.09 seconds (58591.78 events/sec)
INFO:__main__:Inserted batch 65 with 5000 events in 0.08 seconds (60133.05 events/sec)
INFO:__main__:Inserted batch 66 with 5000 events in 0.08 seconds (59571.92 events/sec)
INFO:__main__:Inserted batch 67 with 5000 events in 0.08 seconds (60909.54 events/sec)
INFO:__main__:Inserted batch 68 with 5000 events in 0.08 seconds (61372.10 events/sec)
INFO:__main__:Inserted batch 69 with 5000 events in 0.08 seconds (61232.50 events/sec)
INFO:__main__:Inserted batch 70 with 5000 events in 0.08 seconds (62099.38 events/sec)
INFO:__main__:Inserted batch 71 with 5000 events in 0.08 seconds (61392.04 events/sec)
INFO:__main__:Inserted batch 72 with 5000 events in 0.08 seconds (59276.58 events/sec)
INFO:__main__:Inserted batch 73 with 5000 events in 0.09 seconds (57707.76 events/sec)
INFO:__main__:Inserted batch 74 with 5000 events in 0.08 seconds (59384.51 events/sec)
INFO:__main__:Inserted batch 75 with 5000 events in 0.08 seconds (60872.94 events/sec)
INFO:__main__:Inserted batch 76 with 5000 events in 0.08 seconds (61217.13 events/sec)
INFO:__main__:Inserted batch 77 with 5000 events in 0.08 seconds (61129.34 events/sec)
INFO:__main__:Inserted batch 78 with 5000 events in 0.08 seconds (61061.53 events/sec)
INFO:__main__:Inserted batch 79 with 5000 events in 0.08 seconds (60586.46 events/sec)
INFO:__main__:Inserted batch 80 with 5000 events in 0.08 seconds (59071.04 events/sec)
INFO:__main__:Inserted batch 81 with 5000 events in 0.08 seconds (60347.79 events/sec)
INFO:__main__:Inserted batch 82 with 5000 events in 0.08 seconds (59875.01 events/sec)
INFO:__main__:Inserted batch 83 with 5000 events in 0.09 seconds (57740.65 events/sec)
INFO:__main__:Inserted batch 84 with 5000 events in 0.08 seconds (60582.44 events/sec)
INFO:__main__:Inserted batch 85 with 5000 events in 0.08 seconds (60833.92 events/sec)
INFO:__main__:Inserted batch 86 with 5000 events in 0.09 seconds (58281.74 events/sec)
INFO:__main__:Inserted batch 87 with 5000 events in 0.08 seconds (60078.09 events/sec)
INFO:__main__:Inserted batch 88 with 5000 events in 0.09 seconds (58692.13 events/sec)
INFO:__main__:Inserted batch 89 with 5000 events in 0.08 seconds (60910.07 events/sec)
INFO:__main__:Inserted batch 90 with 5000 events in 0.08 seconds (60866.94 events/sec)
INFO:__main__:Inserted batch 91 with 5000 events in 0.08 seconds (60690.26 events/sec)
INFO:__main__:Inserted batch 92 with 5000 events in 0.08 seconds (59360.47 events/sec)
INFO:__main__:Inserted batch 93 with 5000 events in 0.08 seconds (60913.79 events/sec)
INFO:__main__:Inserted batch 94 with 5000 events in 0.10 seconds (47876.36 events/sec)
INFO:__main__:Inserted batch 95 with 5000 events in 0.08 seconds (61115.09 events/sec)
INFO:__main__:Inserted batch 96 with 5000 events in 0.08 seconds (60889.91 events/sec)
INFO:__main__:Inserted batch 97 with 5000 events in 0.08 seconds (59516.30 events/sec)
INFO:__main__:Inserted batch 98 with 5000 events in 0.09 seconds (52657.99 events/sec)
INFO:__main__:Inserted batch 99 with 5000 events in 0.08 seconds (59015.85 events/sec)
INFO:__main__:Inserted batch 100 with 5000 events in 0.08 seconds (59500.09 events/sec)
INFO:__main__:Inserted batch 101 with 5000 events in 0.08 seconds (59718.83 events/sec)
INFO:__main__:Inserted batch 102 with 5000 events in 0.09 seconds (58250.34 events/sec)
INFO:__main__:Inserted batch 103 with 5000 events in 0.08 seconds (61008.41 events/sec)
INFO:__main__:Inserted batch 104 with 5000 events in 0.08 seconds (60988.36 events/sec)
INFO:__main__:Inserted batch 105 with 5000 events in 0.08 seconds (61100.13 events/sec)
INFO:__main__:Inserted batch 106 with 5000 events in 0.08 seconds (60659.19 events/sec)
INFO:__main__:Inserted batch 107 with 5000 events in 0.08 seconds (61433.94 events/sec)
INFO:__main__:Inserted batch 108 with 5000 events in 0.08 seconds (60162.89 events/sec)
INFO:__main__:Inserted batch 109 with 5000 events in 0.08 seconds (60524.57 events/sec)
INFO:__main__:Inserted batch 110 with 5000 events in 0.09 seconds (58618.63 events/sec)
INFO:__main__:Inserted batch 111 with 5000 events in 0.09 seconds (56181.74 events/sec)
INFO:__main__:Inserted batch 112 with 5000 events in 0.08 seconds (60709.06 events/sec)
INFO:__main__:Inserted batch 113 with 5000 events in 0.08 seconds (60753.91 events/sec)
INFO:__main__:Inserted batch 114 with 5000 events in 0.08 seconds (59026.65 events/sec)
INFO:__main__:Inserted batch 115 with 5000 events in 0.09 seconds (58517.39 events/sec)
INFO:__main__:Inserted batch 116 with 5000 events in 0.08 seconds (61356.83 events/sec)
INFO:__main__:Inserted batch 117 with 5000 events in 0.08 seconds (61392.04 events/sec)
INFO:__main__:Inserted batch 118 with 5000 events in 0.08 seconds (61692.01 events/sec)
INFO:__main__:Inserted batch 119 with 5000 events in 0.09 seconds (57121.47 events/sec)
INFO:__main__:Inserted batch 120 with 5000 events in 0.08 seconds (60300.24 events/sec)
INFO:__main__:Inserted batch 121 with 5000 events in 0.08 seconds (59299.37 events/sec)
INFO:__main__:Inserted batch 122 with 5000 events in 0.08 seconds (62718.19 events/sec)
INFO:__main__:Inserted batch 123 with 5000 events in 0.08 seconds (61665.16 events/sec)
INFO:__main__:Inserted batch 124 with 5000 events in 0.08 seconds (60483.02 events/sec)
INFO:__main__:Inserted batch 125 with 5000 events in 0.08 seconds (60815.75 events/sec)
INFO:__main__:Inserted batch 126 with 5000 events in 0.08 seconds (62324.35 events/sec)
INFO:__main__:Inserted batch 127 with 5000 events in 0.08 seconds (62301.57 events/sec)
INFO:__main__:Inserted batch 128 with 5000 events in 0.08 seconds (59990.79 events/sec)
INFO:__main__:Inserted batch 129 with 5000 events in 0.08 seconds (61615.34 events/sec)
INFO:__main__:Inserted batch 130 with 5000 events in 0.09 seconds (58776.68 events/sec)
INFO:__main__:Inserted batch 131 with 5000 events in 0.08 seconds (60542.04 events/sec)
INFO:__main__:Inserted batch 132 with 5000 events in 0.09 seconds (58653.72 events/sec)
INFO:__main__:Inserted batch 133 with 5000 events in 0.08 seconds (60618.51 events/sec)
INFO:__main__:Inserted batch 134 with 5000 events in 0.08 seconds (62221.17 events/sec)
INFO:__main__:Inserted batch 135 with 5000 events in 0.08 seconds (60850.86 events/sec)
INFO:__main__:Inserted batch 136 with 5000 events in 0.08 seconds (60431.26 events/sec)
INFO:__main__:Inserted batch 137 with 5000 events in 0.08 seconds (59659.71 events/sec)
INFO:__main__:Inserted batch 138 with 5000 events in 0.08 seconds (61355.94 events/sec)
INFO:__main__:Inserted batch 139 with 5000 events in 0.08 seconds (60513.56 events/sec)
INFO:__main__:Inserted batch 140 with 5000 events in 0.09 seconds (57111.83 events/sec)
INFO:__main__:Inserted batch 141 with 5000 events in 0.08 seconds (60274.94 events/sec)
INFO:__main__:Inserted batch 142 with 5000 events in 0.08 seconds (59620.70 events/sec)
INFO:__main__:Inserted batch 143 with 5000 events in 0.08 seconds (60325.39 events/sec)
INFO:__main__:Inserted batch 144 with 5000 events in 0.08 seconds (61003.62 events/sec)
INFO:__main__:Inserted batch 145 with 5000 events in 0.09 seconds (58234.97 events/sec)
INFO:__main__:Inserted batch 146 with 5000 events in 0.08 seconds (60941.22 events/sec)
INFO:__main__:Inserted batch 147 with 5000 events in 0.08 seconds (59946.72 events/sec)
INFO:__main__:Inserted batch 148 with 5000 events in 0.08 seconds (59868.68 events/sec)
INFO:__main__:Inserted batch 149 with 5000 events in 0.09 seconds (58106.53 events/sec)
INFO:__main__:Inserted batch 150 with 5000 events in 0.08 seconds (59796.30 events/sec)
INFO:__main__:Inserted batch 151 with 5000 events in 0.08 seconds (60683.41 events/sec)
INFO:__main__:Inserted batch 152 with 5000 events in 0.08 seconds (60710.82 events/sec)
INFO:__main__:Inserted batch 153 with 5000 events in 0.08 seconds (59928.39 events/sec)
INFO:__main__:Inserted batch 154 with 5000 events in 0.09 seconds (56718.57 events/sec)
INFO:__main__:Inserted batch 155 with 5000 events in 0.08 seconds (61275.80 events/sec)
INFO:__main__:Inserted batch 156 with 5000 events in 0.08 seconds (60462.45 events/sec)
INFO:__main__:Inserted batch 157 with 5000 events in 0.08 seconds (59124.00 events/sec)
INFO:__main__:Inserted batch 158 with 5000 events in 0.08 seconds (58993.61 events/sec)
INFO:__main__:Inserted batch 159 with 5000 events in 0.09 seconds (58088.03 events/sec)
INFO:__main__:Inserted batch 160 with 5000 events in 0.09 seconds (55837.10 events/sec)
INFO:__main__:Inserted batch 161 with 5000 events in 0.08 seconds (59123.34 events/sec)
INFO:__main__:Inserted batch 162 with 5000 events in 0.08 seconds (59958.03 events/sec)
INFO:__main__:Inserted batch 163 with 5000 events in 0.08 seconds (59272.56 events/sec)
INFO:__main__:Inserted batch 164 with 5000 events in 0.08 seconds (59924.28 events/sec)
INFO:__main__:Inserted batch 165 with 5000 events in 0.08 seconds (60516.36 events/sec)
INFO:__main__:Inserted batch 166 with 5000 events in 0.08 seconds (61169.81 events/sec)
INFO:__main__:Inserted batch 167 with 5000 events in 0.08 seconds (61027.05 events/sec)
INFO:__main__:Inserted batch 168 with 5000 events in 0.09 seconds (57761.48 events/sec)
INFO:__main__:Inserted batch 169 with 5000 events in 0.08 seconds (59602.06 events/sec)
INFO:__main__:Inserted batch 170 with 5000 events in 0.08 seconds (59260.83 events/sec)
INFO:__main__:Inserted batch 171 with 5000 events in 0.08 seconds (61051.57 events/sec)
INFO:__main__:Inserted batch 172 with 5000 events in 0.08 seconds (60056.42 events/sec)
INFO:__main__:Inserted batch 173 with 5000 events in 0.08 seconds (60164.27 events/sec)
INFO:__main__:Inserted batch 174 with 5000 events in 0.08 seconds (59361.65 events/sec)
INFO:__main__:Inserted batch 175 with 5000 events in 0.09 seconds (57890.96 events/sec)
INFO:__main__:Inserted batch 176 with 5000 events in 0.09 seconds (57899.43 events/sec)
INFO:__main__:Inserted batch 177 with 5000 events in 0.08 seconds (61684.03 events/sec)
INFO:__main__:Inserted batch 178 with 5000 events in 0.09 seconds (56205.23 events/sec)
INFO:__main__:Inserted batch 179 with 5000 events in 0.08 seconds (60313.42 events/sec)
INFO:__main__:Inserted batch 180 with 5000 events in 0.09 seconds (58260.21 events/sec)
INFO:__main__:Inserted batch 181 with 5000 events in 0.08 seconds (60499.42 events/sec)
INFO:__main__:Inserted batch 182 with 5000 events in 0.09 seconds (55690.89 events/sec)
INFO:__main__:Inserted batch 183 with 5000 events in 0.09 seconds (56410.54 events/sec)
INFO:__main__:Inserted batch 184 with 5000 events in 0.09 seconds (57906.14 events/sec)
INFO:__main__:Inserted batch 185 with 5000 events in 0.09 seconds (56736.98 events/sec)
INFO:__main__:Inserted batch 186 with 5000 events in 0.08 seconds (59127.17 events/sec)
INFO:__main__:Inserted batch 187 with 5000 events in 0.09 seconds (56599.32 events/sec)
INFO:__main__:Inserted batch 188 with 5000 events in 0.08 seconds (60208.49 events/sec)
INFO:__main__:Inserted batch 189 with 5000 events in 0.08 seconds (59523.73 events/sec)
INFO:__main__:Inserted batch 190 with 5000 events in 0.09 seconds (58641.42 events/sec)
INFO:__main__:Inserted batch 191 with 5000 events in 0.08 seconds (61436.46 events/sec)
INFO:__main__:Inserted batch 192 with 5000 events in 0.09 seconds (58496.50 events/sec)
INFO:__main__:Inserted batch 193 with 5000 events in 0.09 seconds (56366.72 events/sec)
INFO:__main__:Inserted batch 194 with 5000 events in 0.08 seconds (60061.23 events/sec)
INFO:__main__:Inserted batch 195 with 5000 events in 0.09 seconds (57494.97 events/sec)
INFO:__main__:Inserted batch 196 with 5000 events in 0.09 seconds (55713.23 events/sec)
INFO:__main__:Inserted batch 197 with 5000 events in 0.09 seconds (54448.85 events/sec)
INFO:__main__:Inserted batch 198 with 5000 events in 0.08 seconds (62113.91 events/sec)
INFO:__main__:Inserted batch 199 with 5000 events in 0.09 seconds (58062.45 events/sec)
INFO:__main__:Inserted batch 200 with 5000 events in 0.09 seconds (58331.02 events/sec)
INFO:__main__:Total: Inserted 1000000 events in 16.91 seconds (59148.66 events/sec)
INFO:__main__:Successfully inserted 1000000 events in 16.91 seconds写入前后的磁盘空间占用
写入前
[root@localhost ~]# du -sh /data/clickhouse/
70M /data/clickhouse/
[root@localhost ~]#写入后
[root@localhost clickhouse-sample]# du -sh /data/clickhouse/
333M /data/clickhouse/
[root@localhost clickhouse-sample]#写入后的内存
CONTAINER ID NAME CPU % MEM USAGE / LIMIT MEM % NET I/O BLOCK I/O PIDS
0bab25db35bf clickhouse-server 2.89% 851.8MiB / 34.63GiB 2.40% 180MB / 987kB 102MB / 819kB 691简单查看写入的数据
0bab25db35bf :) select * from cmo_events limit 20;
SELECT *
FROM cmo_events
LIMIT 20
Query id: 37fb0dea-dae6-4d27-8d79-62cba659b7e0
┌─event_id─────────────────────────────┬─session_uuid─────────────────────────┬──────────event_time─┬─event_type──────────────┬─side─┬─unit_type─────────────────┬─unit_name────────────────────────┬─weapon_system─────────┬─target_type────┬─target_name────────────┬─damage_points─┬───────location_lat─┬───────location_lon─┬─description──────────────────────────────────────────────────────────────────────────────┐
1. │ 8849c00f-cd82-46b2-ab9d-ec9c9de87890 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ COMMUNICATION │ BLUE │ Zumwalt-class DDG │ Blue Zumwalt-class DDG #10 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 28.10526232009602 │ 121.95699352544986 │ Blue Zumwalt-class DDG #10 communication │
2. │ a872a0a9-1252-4813-ab8c-4cff1f401f8d │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ LANDING │ RED │ Borei-class SSBN │ Red Borei-class SSBN #6 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 30.696694625440067 │ 121.29328734609689 │ Red Borei-class SSBN #6 landed at base │
3. │ 1ad76e63-006f-4e91-83ab-a700828cd82a │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ DAMAGE_ASSESSMENT │ BLUE │ Arleigh Burke-class DDG │ Blue Arleigh Burke-class DDG #1 │ ᴺᵁᴸᴸ │ GROUND_VEHICLE │ RED GROUND_VEHICLE #14 │ 87 │ 26.261531960506662 │ 121.59999706737347 │ Blue Arleigh Burke-class DDG #1 damage assessment │
4. │ bdcd36a4-3b36-481c-89e6-b2c422aba5b0 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ WEAPON_HIT │ BLUE │ F-35 Lightning II │ Blue F-35 Lightning II #3 │ Torpedo 53-65 │ INFRASTRUCTURE │ RED INFRASTRUCTURE #3 │ 65 │ 33.112945272285366 │ 118.45447685467576 │ Torpedo 53-65 from Blue F-35 Lightning II #3 hit RED INFRASTRUCTURE #3 causing 65 damage │
5. │ 0e9a9c2f-7bcb-481d-8ae3-e869c934615f │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ SENSOR_ACTIVATION │ RED │ Borei-class SSBN │ Red Borei-class SSBN #5 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 33.10888852294174 │ 121.66802049225652 │ Red Borei-class SSBN #5 sensor activation │
6. │ 7d90b6fa-8775-449e-a9e8-becb05c63af1 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ DETECTION │ RED │ S-400 Triumf │ Red S-400 Triumf #8 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 31.324694206355506 │ 119.61740383324297 │ Red S-400 Triumf #8 detected unknown contact │
7. │ f054c911-b909-485b-bf58-ec389c68f279 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ LANDING │ RED │ F-22A Raptor │ Red F-22A Raptor #7 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 28.983355606808068 │ 118.87768468760133 │ Red F-22A Raptor #7 landed at base │
8. │ 3caa95bf-0ece-4a07-b115-70ba1fc248b5 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ SENSOR_ACTIVATION │ RED │ Yasen-class SSN │ Red Yasen-class SSN #11 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 27.212144356311303 │ 123.75301138525788 │ Red Yasen-class SSN #11 sensor activation │
9. │ c5c1e9b6-0eed-4146-b7fa-1fa3378e1b07 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ WEAPON_MISS │ BLUE │ Virginia-class SSN │ Blue Virginia-class SSN #10 │ Kalibr Cruise Missile │ GROUND_VEHICLE │ RED GROUND_VEHICLE #11 │ ᴺᵁᴸᴸ │ 34.917367440747384 │ 118.09356032516219 │ Kalibr Cruise Missile from Blue Virginia-class SSN #10 missed RED GROUND_VEHICLE #11 │
10. │ d9728362-f068-4a40-874c-f73ec4f473cf │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ FUEL_STATUS │ BLUE │ Arleigh Burke-class DDG │ Blue Arleigh Burke-class DDG #12 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 27.645216291227896 │ 120.54220071943406 │ Blue Arleigh Burke-class DDG #12 fuel status │
11. │ 4a8aff13-84df-47a8-8f5f-806182f185c1 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ MISSION_UPDATE │ RED │ Su-57 Felon │ Red Su-57 Felon #7 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 33.79360033184239 │ 122.92032627373821 │ Red Su-57 Felon #7 mission update │
12. │ ea4d9ae7-856c-4e2d-ab9e-a11a2c6acf1f │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ MISSION_UPDATE │ RED │ Su-57 Felon │ Red Su-57 Felon #2 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 31.67506881472707 │ 115.28273231630332 │ Red Su-57 Felon #2 mission update │
13. │ 0fee2c66-1cf9-40e9-b482-89c2b1332a63 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ SENSOR_ACTIVATION │ BLUE │ M1A2 Abrams │ Blue M1A2 Abrams #9 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 27.01940220507797 │ 116.35083989530673 │ Blue M1A2 Abrams #9 sensor activation │
14. │ e3441e5d-a55e-4339-8d2a-19a5cb88c072 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ WEAPON_LAUNCH │ RED │ Kirov-class Battlecruiser │ Red Kirov-class Battlecruiser #2 │ R-73 Archer │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 25.57873263183426 │ 116.61851532134524 │ Red Kirov-class Battlecruiser #2 launched R-73 Archer │
15. │ 9599d2ca-3206-45a9-8229-f1213dce2930 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ DETECTION │ RED │ J-20 Mighty Dragon │ Red J-20 Mighty Dragon #4 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 34.4541596455493 │ 117.62987749184556 │ Red J-20 Mighty Dragon #4 detected unknown contact │
16. │ b8a471e5-217b-4ee2-83d6-dda9e652f3af │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ LANDING │ BLUE │ M1A2 Abrams │ Blue M1A2 Abrams #6 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 33.169548654507565 │ 118.58063252886755 │ Blue M1A2 Abrams #6 landed at base │
17. │ b10edb56-3180-42ae-bd27-5c714819982a │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ DETECTION │ BLUE │ Patriot SAM │ Blue Patriot SAM #11 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 32.81706047928581 │ 116.25912274827049 │ Blue Patriot SAM #11 detected unknown contact │
18. │ 46795b70-2d03-4b05-828e-41a0a64475ed │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ MISSION_UPDATE │ RED │ Su-57 Felon │ Red Su-57 Felon #1 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 30.98647515114664 │ 123.10269312841808 │ Red Su-57 Felon #1 mission update │
19. │ 458402f7-a738-43e0-b303-7326ff2e900c │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ COUNTERMEASURE_DEPLOYED │ RED │ Borei-class SSBN │ Red Borei-class SSBN #1 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 29.84279445458373 │ 119.76491994064281 │ Red Borei-class SSBN #1 countermeasure deployed │
20. │ 389aca11-e7b8-431a-ab42-7e54306ecc92 │ 09cfee87-3b10-41cb-a995-9bb351944997 │ 2025-09-20 10:24:30 │ MISSION_UPDATE │ BLUE │ F-35 Lightning II │ Blue F-35 Lightning II #2 │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ ᴺᵁᴸᴸ │ 33.58536620950474 │ 124.293345407109 │ Blue F-35 Lightning II #2 mission update │
└──────────────────────────────────────┴──────────────────────────────────────┴─────────────────────┴─────────────────────────┴──────┴───────────────────────────┴──────────────────────────────────┴───────────────────────┴────────────────┴────────────────────────┴───────────────┴────────────────────┴────────────────────┴──────────────────────────────────────────────────────────────────────────────────────────┘
20 rows in set. Elapsed: 0.018 sec.实时评估十项典型指标
analyze_cmo_data.py
from clickhouse_driver import Client
import pandas as pd
from tabulate import tabulate
import matplotlib.pyplot as plt
import matplotlib
matplotlib.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
matplotlib.rcParams['axes.unicode_minus'] = False # 用来正常显示负号
# ClickHouse连接配置
CH_HOST = 'localhost'
CH_PORT = 32772 # 默认原生TCP端口
CH_USER = 'default'
CH_PASSWORD = 'nQczEy2EWJ7w' # 如果设置了密码，请填写
CH_DATABASE = 'default'
CH_TABLE = 'cmo_events'
def create_clickhouse_client():
"""创建ClickHouse客户端连接"""
try:
client = Client(
host=CH_HOST,
port=CH_PORT,
user=CH_USER,
password=CH_PASSWORD,
database=CH_DATABASE
)
print("成功连接到ClickHouse数据库")
return client
except Exception as e:
print(f"连接ClickHouse失败: {e}")
return None
def execute_query(client, query):
"""执行查询并返回结果"""
try:
result = client.execute(query)
return result
except Exception as e:
print(f"查询执行失败: {e}")
return None
def print_chinese_report(title, headers, data):
"""打印中文报表"""
print(f"\n{'='*60}")
print(f"{title:^60}")
print(f"{'='*60}")
print(tabulate(data, headers=headers, tablefmt='grid', numalign="right"))
print(f"{'='*60}")
def analyze_cmo_data(client):
"""执行CMO推演数据分析"""
# 1. 总体统计
print("正在进行CMO推演数据分析...")
total_events = execute_query(client, f"SELECT COUNT(*) FROM {CH_TABLE}")[0][0]
session_uuid = execute_query(client, f"SELECT session_uuid FROM {CH_TABLE} LIMIT 1")[0][0]
print(f"\n推演场次UUID: {session_uuid}")
print(f"总事件数量: {total_events:,}")
# 2. 红蓝双方事件数量对比
side_stats = execute_query(client, f"""
SELECT side, COUNT(*) as count,
COUNT(*) * 100.0 / (SELECT COUNT(*) FROM {CH_TABLE}) as percentage
FROM {CH_TABLE}
GROUP BY side
ORDER BY count DESC
""")
print_chinese_report(
"红蓝双方事件数量对比",
["阵营", "事件数量", "占比(%)"],
side_stats
)
# 3. 事件类型分布
event_type_stats = execute_query(client, f"""
SELECT event_type, COUNT(*) as count,
COUNT(*) * 100.0 / (SELECT COUNT(*) FROM {CH_TABLE}) as percentage
FROM {CH_TABLE}
GROUP BY event_type
ORDER BY count DESC
LIMIT 10
""")
print_chinese_report(
"事件类型分布TOP10",
["事件类型", "发生次数", "占比(%)"],
event_type_stats
)
# 4. 最活跃的作战单位
active_units = execute_query(client, f"""
SELECT unit_name, COUNT(*) as event_count
FROM {CH_TABLE}
GROUP BY unit_name
ORDER BY event_count DESC
LIMIT 10
""")
print_chinese_report(
"最活跃的作战单位TOP10",
["单位名称", "事件数量"],
active_units
)
# 5. 武器使用统计
weapon_stats = execute_query(client, f"""
SELECT weapon_system, COUNT(*) as usage_count
FROM {CH_TABLE}
WHERE weapon_system IS NOT NULL
GROUP BY weapon_system
ORDER BY usage_count DESC
LIMIT 10
""")
print_chinese_report(
"武器使用统计TOP10",
["武器系统", "使用次数"],
weapon_stats
)
# 6. 命中与伤害统计
hit_stats = execute_query(client, f"""
SELECT
COUNTIf(event_type = 'WEAPON_HIT') as hit_count,
COUNTIf(event_type = 'WEAPON_MISS') as miss_count,
SUMIf(damage_points, event_type = 'WEAPON_HIT') as total_damage,
AVGIf(damage_points, event_type = 'WEAPON_HIT') as avg_damage
FROM {CH_TABLE}
""")
hit_ratio = hit_stats[0][0] / (hit_stats[0][0] + hit_stats[0][1]) * 100 if (hit_stats[0][0] + hit_stats[0][1]) > 0 else 0
print_chinese_report(
"武器命中与伤害统计",
["命中次数", "未命中次数", "命中率(%)", "总伤害", "平均伤害"],
[[hit_stats[0][0], hit_stats[0][1], f"{hit_ratio:.2f}", hit_stats[0][2] or 0, f"{(hit_stats[0][3] or 0):.2f}"]]
)
# 7. 时间线分析 - 事件频率
timeline_stats = execute_query(client, f"""
SELECT
toStartOfMinute(event_time) as minute,
COUNT(*) as events_per_minute
FROM {CH_TABLE}
GROUP BY minute
ORDER BY minute
""")
# 转换为DataFrame以便分析
timeline_df = pd.DataFrame(timeline_stats, columns=['minute', 'events_per_minute'])
print_chinese_report(
"时间线分析 - 事件频率",
["统计项", "值"],
[
["推演开始时间", timeline_df['minute'].min()],
["推演结束时间", timeline_df['minute'].max()],
["推演持续时间(分钟)", (timeline_df['minute'].max() - timeline_df['minute'].min()).total_seconds() / 60],
["每分钟平均事件数", timeline_df['events_per_minute'].mean()],
["每分钟最大事件数", timeline_df['events_per_minute'].max()],
["每分钟最小事件数", timeline_df['events_per_minute'].min()]
]
)
# 8. 目标类型分析
target_stats = execute_query(client, f"""
SELECT target_type, COUNT(*) as count
FROM {CH_TABLE}
WHERE target_type IS NOT NULL
GROUP BY target_type
ORDER BY count DESC
LIMIT 10
""")
print_chinese_report(
"目标类型分析TOP10",
["目标类型", "被攻击次数"],
target_stats
)
# 9. 地理分布分析
geo_stats = execute_query(client, f"""
SELECT
round(location_lat, 1) as lat_zone,
round(location_lon, 1) as lon_zone,
COUNT(*) as event_count
FROM {CH_TABLE}
GROUP BY lat_zone, lon_zone
ORDER BY event_count DESC
LIMIT 10
""")
print_chinese_report(
"热点区域分析TOP10",
["纬度区域", "经度区域", "事件数量"],
geo_stats
)
# 10. 推演阶段分析
phase_analysis = execute_query(client, f"""
WITH time_buckets AS (
SELECT
event_time,
CASE
WHEN event_time < (SELECT MIN(event_time) FROM {CH_TABLE}) + INTERVAL 2 HOUR THEN '初期阶段'
WHEN event_time < (SELECT MIN(event_time) FROM {CH_TABLE}) + INTERVAL 4 HOUR THEN '中期阶段'
ELSE '后期阶段'
END as phase
FROM {CH_TABLE}
)
SELECT
phase,
COUNT(*) as event_count,
COUNTIf(event_type = 'WEAPON_LAUNCH') as weapon_launches,
COUNTIf(event_type = 'WEAPON_HIT') as weapon_hits
FROM time_buckets
JOIN {CH_TABLE} USING (event_time)
GROUP BY phase
ORDER BY phase
""")
print_chinese_report(
"推演阶段分析",
["阶段", "事件总数", "武器发射次数", "命中次数"],
phase_analysis
)
# 额外分析：武器效率
weapon_efficiency = execute_query(client, f"""
SELECT
weapon_system,
COUNT(*) as total_uses,
COUNTIf(event_type = 'WEAPON_HIT') as hits,
COUNTIf(event_type = 'WEAPON_MISS') as misses,
CASE
WHEN COUNTIf(event_type IN ('WEAPON_HIT', 'WEAPON_MISS')) > 0
THEN COUNTIf(event_type = 'WEAPON_HIT') * 100.0 / COUNTIf(event_type IN ('WEAPON_HIT', 'WEAPON_MISS'))
ELSE 0
END as hit_rate
FROM {CH_TABLE}
WHERE weapon_system IS NOT NULL
GROUP BY weapon_system
HAVING total_uses >= 10
ORDER BY hit_rate DESC
LIMIT 10
""")
print_chinese_report(
"武器效率分析TOP10 (至少使用10次)",
["武器系统", "使用次数", "命中次数", "未命中次数", "命中率(%)"],
weapon_efficiency
)
def main():
"""主函数"""
print("CMO推演数据分析程序")
print("开始连接ClickHouse数据库...")
client = create_clickhouse_client()
if client is None:
return
try:
analyze_cmo_data(client)
print("\n分析完成！")
except Exception as e:
print(f"分析过程中出现错误: {e}")
finally:
client.disconnect()
if __name__ == "__main__":
# 安装依赖: pip install clickhouse-driver pandas tabulate matplotlib
main()运行效果
[root@localhost clickhouse-sample]# time python3 analyze_cmo_data.py
CMO推演数据分析程序
开始连接ClickHouse数据库...
成功连接到ClickHouse数据库
正在进行CMO推演数据分析...
推演场次UUID: 09cfee87-3b10-41cb-a995-9bb351944997
总事件数量: 1,000,000
============================================================
红蓝双方事件数量对比
============================================================
+--------+------------+-----------+
| 阵营 | 事件数量 | 占比(%) |
+========+============+===========+
| RED | 500045 | 50.0045 |
+--------+------------+-----------+
| BLUE | 499955 | 49.9955 |
+--------+------------+-----------+
============================================================
============================================================
事件类型分布TOP10
============================================================
+-------------------------+------------+-----------+
| 事件类型 | 发生次数 | 占比(%) |
+=========================+============+===========+
| WEAPON_HIT | 83847 | 8.3847 |
+-------------------------+------------+-----------+
| DETECTION | 83808 | 8.3808 |
+-------------------------+------------+-----------+
| LANDING | 83490 | 8.349 |
+-------------------------+------------+-----------+
| WEAPON_LAUNCH | 83438 | 8.3438 |
+-------------------------+------------+-----------+
| COUNTERMEASURE_DEPLOYED | 83413 | 8.3413 |
+-------------------------+------------+-----------+
| FUEL_STATUS | 83361 | 8.3361 |
+-------------------------+------------+-----------+
| WEAPON_MISS | 83357 | 8.3357 |
+-------------------------+------------+-----------+
| TAKEOFF | 83227 | 8.3227 |
+-------------------------+------------+-----------+
| SENSOR_ACTIVATION | 83173 | 8.3173 |
+-------------------------+------------+-----------+
| DAMAGE_ASSESSMENT | 83071 | 8.3071 |
+-------------------------+------------+-----------+
============================================================
============================================================
最活跃的作战单位TOP10
============================================================
+----------------------------------+------------+
| 单位名称 | 事件数量 |
+==================================+============+
| Red Kirov-class Battlecruiser #4 | 5406 |
+----------------------------------+------------+
| Red S-400 Triumf #2 | 5379 |
+----------------------------------+------------+
| Blue M1A2 Abrams #6 | 5367 |
+----------------------------------+------------+
| Red J-20 Mighty Dragon #10 | 5350 |
+----------------------------------+------------+
| Red F-22A Raptor #10 | 5348 |
+----------------------------------+------------+
| Blue Trident II D5 SLBM #11 | 5329 |
+----------------------------------+------------+
| Red Kirov-class Battlecruiser #1 | 5325 |
+----------------------------------+------------+
| Blue F-35 Lightning II #7 | 5323 |
+----------------------------------+------------+
| Blue F-35 Lightning II #9 | 5321 |
+----------------------------------+------------+
| Blue Patriot SAM #6 | 5319 |
+----------------------------------+------------+
============================================================
============================================================
武器使用统计TOP10
============================================================
+-------------------------+------------+
| 武器系统 | 使用次数 |
+=========================+============+
| R-73 Archer | 22971 |
+-------------------------+------------+
| AIM-120D AMRAAM | 22889 |
+-------------------------+------------+
| Torpedo MK-48 | 22856 |
+-------------------------+------------+
| Tomahawk Cruise Missile | 22856 |
+-------------------------+------------+
| Kalibr Cruise Missile | 22829 |
+-------------------------+------------+
| R-77 Adder | 22803 |
+-------------------------+------------+
| Torpedo 53-65 | 22779 |
+-------------------------+------------+
| PL-15 | 22777 |
+-------------------------+------------+
| AIM-9X Sidewinder | 22739 |
+-------------------------+------------+
| DF-21D ASBM | 22611 |
+-------------------------+------------+
============================================================
============================================================
武器命中与伤害统计
============================================================
+------------+--------------+-------------+----------+------------+
| 命中次数 | 未命中次数 | 命中率(%) | 总伤害 | 平均伤害 |
+============+==============+=============+==========+============+
| 83847 | 83357 | 50.15 | 4606780 | 54.94 |
+------------+--------------+-------------+----------+------------+
============================================================
============================================================
时间线分析 - 事件频率
============================================================
+--------------------+---------------------+
| 统计项 | 值 |
+====================+=====================+
| 推演开始时间 | 2025-09-20 10:24:00 |
+--------------------+---------------------+
| 推演结束时间 | 2025-09-20 16:24:00 |
+--------------------+---------------------+
| 推演持续时间(分钟) | 360.0 |
+--------------------+---------------------+
| 每分钟平均事件数 | 2770.083102493075 |
+--------------------+---------------------+
| 每分钟最大事件数 | 2915 |
+--------------------+---------------------+
| 每分钟最小事件数 | 1421 |
+--------------------+---------------------+
============================================================
============================================================
目标类型分析TOP10
============================================================
+----------------+--------------+
| 目标类型 | 被攻击次数 |
+================+==============+
| SUBMAIRNE | 27956 |
+----------------+--------------+
| AIRCRAFT | 27882 |
+----------------+--------------+
| RADAR_SITE | 27872 |
+----------------+--------------+
| AIRBASE | 27858 |
+----------------+--------------+
| INFRASTRUCTURE | 27856 |
+----------------+--------------+
| GROUND_VEHICLE | 27775 |
+----------------+--------------+
| WARSHIP | 27759 |
+----------------+--------------+
| COMMAND_CENTER | 27728 |
+----------------+--------------+
| SAM_SITE | 27589 |
+----------------+--------------+
============================================================
============================================================
热点区域分析TOP10
============================================================
+------------+------------+------------+
| 纬度区域 | 经度区域 | 事件数量 |
+============+============+============+
| 33.3 | 118.3 | 144 |
+------------+------------+------------+
| 32.5 | 119.6 | 140 |
+------------+------------+------------+
| 29.9 | 121.1 | 136 |
+------------+------------+------------+
| 26.4 | 121.7 | 136 |
+------------+------------+------------+
| 33.9 | 119.3 | 136 |
+------------+------------+------------+
| 31.1 | 116.7 | 135 |
+------------+------------+------------+
| 27.8 | 115.5 | 135 |
+------------+------------+------------+
| 33 | 122.5 | 134 |
+------------+------------+------------+
| 30.6 | 118.3 | 132 |
+------------+------------+------------+
| 26.1 | 122.2 | 131 |
+------------+------------+------------+
============================================================
============================================================
推演阶段分析
============================================================
+----------+------------+----------------+------------+
| 阶段 | 事件总数 | 武器发射次数 | 命中次数 |
+==========+============+================+============+
| 中期阶段 | 15747891 | 1322005 | 1322311 |
+----------+------------+----------------+------------+
| 初期阶段 | 15878750 | 1323617 | 1319577 |
+----------+------------+----------------+------------+
| 后期阶段 | 15673439 | 1304067 | 1323536 |
+----------+------------+----------------+------------+
============================================================
============================================================
武器效率分析TOP10 (至少使用10次)
============================================================
+-------------------------+------------+------------+--------------+-------------+
| 武器系统 | 使用次数 | 命中次数 | 未命中次数 | 命中率(%) |
+=========================+============+============+==============+=============+
| Kalibr Cruise Missile | 22829 | 7817 | 7496 | 51.0481 |
+-------------------------+------------+------------+--------------+-------------+
| R-77 Adder | 22803 | 7737 | 7518 | 50.7178 |
+-------------------------+------------+------------+--------------+-------------+
| PL-15 | 22777 | 7648 | 7488 | 50.5285 |
+-------------------------+------------+------------+--------------+-------------+
| Standard Missile-6 | 22532 | 7566 | 7431 | 50.4501 |
+-------------------------+------------+------------+--------------+-------------+
| AIM-120D AMRAAM | 22889 | 7732 | 7637 | 50.3091 |
+-------------------------+------------+------------+--------------+-------------+
| Torpedo MK-48 | 22856 | 7707 | 7637 | 50.2281 |
+-------------------------+------------+------------+--------------+-------------+
| DF-21D ASBM | 22611 | 7536 | 7537 | 49.9967 |
+-------------------------+------------+------------+--------------+-------------+
| R-73 Archer | 22971 | 7586 | 7640 | 49.8227 |
+-------------------------+------------+------------+--------------+-------------+
| Tomahawk Cruise Missile | 22856 | 7572 | 7677 | 49.6557 |
+-------------------------+------------+------------+--------------+-------------+
| Torpedo 53-65 | 22779 | 7561 | 7686 | 49.5901 |
+-------------------------+------------+------------+--------------+-------------+
============================================================
分析完成！
real 0m1.314s
user 0m7.131s
sys 0m0.079s可以看到，clickhouse在100万个事件中做十项统计，耗时不到2s
参考
Install ClickHouse using Docker | ClickHouse Docs
附：一linux time指令时间解释
您观察到的 real、user 和 sys 时间之间的差异是正常的，这反映了程序运行时的不同时间维度：
时间类型解释
-
real (实际时间/墙上时间): 0m1.314s
- 这是程序从开始到结束的实际经过时间
- 包括所有等待时间（如I/O操作、网络请求、系统调度等）
-
user (用户CPU时间): 0m7.131s
- 这是程序在用户模式下执行所花费的CPU时间总和
- 如果程序使用了多线程或多进程，这个时间会是所有线程/进程的CPU时间之和
-
sys (系统CPU时间): 0m0.079s
- 这是程序在内核模式下执行所花费的CPU时间
- 包括系统调用、内核服务等
为什么 user 时间比 real 时间长？
在您的情况下，user 时间 (7.131s) 远大于 real 时间 (1.314s)，这通常表明：
-
多线程/多核利用: 您的程序可能使用了多线程，并且运行在多核CPU上。ClickHouse驱动程序可能在后台使用多个线程来处理查询和数据反序列化。
-
并行处理: 虽然您的代码看起来是单线程的，但底层的ClickHouse驱动程序和Python解释器可能在某些操作上使用了并行处理。
-
CPU密集型操作: 数据分析程序涉及大量的计算和数据处理，这些操作是CPU密集型的，因此会累积较多的用户CPU时间。
-
向量化操作: Pandas和其他数据处理库使用向量化操作，这些操作可以高效利用CPU的多个核心。
具体分析
对于您的CMO数据分析程序：
-
查询执行: 程序向ClickHouse发送了多个查询，这些查询可能在服务器端并行执行。
-
数据处理: 使用Pandas处理查询结果时，Pandas会利用多核CPU进行向量化操作。
-
网络I/O: 虽然网络请求是I/O密集型操作，但数据处理是CPU密集型操作，这导致了user时间的累积。
这是一个好现象
实际上，user 时间远大于 real 时间通常是一个好迹象，因为它表明：
- 您的程序高效地利用了多核CPU
- 计算任务被有效地并行化
- CPU资源得到了充分利用
- 程序性能良好，没有因为I/O等待而浪费时间
如何验证
如果您想确认是否是并行处理导致了这种现象，可以：
- 使用单线程运行程序（设置环境变量
OMP_NUM_THREADS=1） - 使用性能分析工具如
cProfile来分析程序的时间分布 - 检查ClickHouse驱动的文档，了解它是否默认使用多线程
总之，您观察到的时间差异是正常的，表明您的程序正在高效地利用系统资源。
附二：ClickHouse是否能支撑同时100场推演的数据写入？
根据您的测试数据和ClickHouse的架构特性分析，答案是肯定的，但需要满足以下条件：
1. 单节点性能评估
- 测试表现：单节点写入100万事件用时16.91秒（59,148事件/秒），且内存占用仅851.8MiB，磁盘占用333MB。
- 扩展性测试推演 ：
若每场推演数据量相同（100万事件），100场总数据量为1亿事件。单节点理论耗时约 28分钟 （100/59,148≈0.0017秒/事件 ×1e8=16910秒）。
但实际并发写入会因资源竞争显著降低性能。
2. 并发写入瓶颈分析
- 磁盘IO：MergeTree引擎写入时使用顺序IO，但100个并发会话可能造成磁盘随机IO增加。
- CPU限制：每条事件生成包含随机数、UUID、字符串操作，CPU密集型操作。
- 锁竞争 ：MergeTree表引擎在高并发写入时可能触发写锁。
3. 分布式架构方案（推荐）
建议采用以下架构实现100场并发写入：
写入客户端 Distributed Table MergeTree分片1 MergeTree分片2 ... MergeTree分片N
关键配置：
sql
-- 创建分布式表（需在集群配置文件中定义分片）
CREATE TABLE cmo_events_distributed AS cmo_events
ENGINE = Distributed('cluster_name', default, cmo_events, session_uuid)Python代码调整：
python
# 修改插入语句为分布式表
insert_query = "INSERT INTO cmo_events_distributed (...) VALUES"优势：
- 线性扩展：通过增加分片数量提升总写入吞吐量
- 负载均衡 ：按
session_uuid哈希分布到不同节点 - 容错能力：单节点故障不影响整体写入
二、腾讯云ClickHouse是否支持在线扩容？
完全支持，其扩容机制具有以下特点：
1. 在线扩容能力
- 无感知扩容 ：通过控制台/API添加节点组，扩容过程中无需停机
- 自动数据再平衡 ：新增节点后自动触发ReplicatedMergeTree数据复制
- 集群拓扑自动更新：节点组扩容后立即参与分布式查询和写入
2. 扩容操作步骤（腾讯云控制台）
登录控制台 选择集群 点击扩容 选择节点组 设置新节点数量 确认配置 自动扩容完成
3. 扩容后性能提升
- 写入吞吐量：近似线性增长（每新增1节点可提升5.9万+事件/秒）
- 查询性能：分布式查询并行加速
- 存储容量：通过增加节点扩展总存储空间
三、生产环境建议配置
| 维度 | 单机方案 | 腾讯云分布式方案 |
|---|---|---|
| 节点数量 | 1 | ≥3 |
| 数据分布 | 单表写入 | Distributed表 + ReplicatedMergeTree |
| 扩容方式 | 手动迁移数据 | 控制台点击扩容 |
| 写入压力 | ≤10万事件/秒 | 按需扩展至百万级/秒 |
| 数据一致性 | 本地MergeTree | ZooKeeper协调 |
| 运维复杂度 | 低 | 云服务托管 |
| 适用场景 | 单场景/小规模测试 | 多推演并发/生产环境 |
四、性能优化补充建议
-
批量写入优化
python# 增大批量大小（测试10000~50000条/批） batch_size = 50000 # 根据内存调整 -
压缩数据
sql-- 启用LZ4压缩 CREATE TABLE cmo_events ENGINE = MergeTree() ORDER BY ... SETTINGS compression_codec = 'LZ4' -
分区策略
sql-- 按session_uuid分区（适合按场次查询） ENGINE = MergeTree() PARTITION BY session_uuid ORDER BY (session_uuid, event_time)
五、总结
- 本地ClickHouse：通过分布式引擎可支撑100场并发推演，但需手动部署集群
- 腾讯云ClickHouse ：
- 支持在线扩容（控制台操作即可）
- 建议使用分布式+ReplicatedMergeTree架构
- 推演场次按
session_uuid分布到不同分片
🚀 实测建议：腾讯云ClickHouse单节点写入能力可达8万10万事件/秒**（取决于数据复杂度），100场并发推演可通过部署**510个节点轻松实现
附三：分片
ClickHouse 的 Distributed 表引擎是分布式查询的核心机制，其本质是一个 逻辑表 ，不存储数据 ，而是作为底层分片表的 代理层，用于：
- 将写入请求分发到所有分片节点
- 将读取请求聚合所有分片节点的数据
一、Distributed 表的本质
1. 不存储数据，只路由
-
cmo_events_distributed表本身不存储任何数据 -
它只是告诉 ClickHouse：
"当我收到写入请求时，请自动把数据分发到所有分片节点的
cmo_events表中"
2. 分片规则由你定义
- 你可以指定分片规则（如哈希、随机、固定分片等）
- 示例：按
session_uuid哈希分布
sql
CREATE TABLE cmo_events_distributed (
...
) ENGINE = Distributed('cluster_name', default, cmo_events, session_uuid)Distributed表会根据session_uuid的哈希值，将数据分发到对应节点的本地cmo_events表
二、Distributed 表 vs 本地表
| 类型 | 本地表（MergeTree） | Distributed 表 |
|---|---|---|
| 数据存储 | 是（存储在单个节点） | 否（不存储数据） |
| 写入操作 | 直接写入本地 | 自动分发到所有分片 |
| 读取操作 | 仅读取本地数据 | 自动聚合所有分片数据 |
| 容错能力 | 单节点故障数据丢失 | 自动读取所有分片 + Replicated表保障 |
| 运维复杂度 | 低 | 需配置集群（或依赖云服务） |
| 典型用途 | 数据存储 | 分布式写入/查询入口 |
三、Distributed 表的创建逻辑
1. 本地表（真实存储）
sql
-- 在每个节点上创建
CREATE TABLE cmo_events (
event_id UUID,
session_uuid UUID,
event_time DateTime,
...
) ENGINE = MergeTree()
ORDER BY (session_uuid, event_time)2. Distributed 表（逻辑代理）
sql
-- 在集群任意节点创建
CREATE TABLE cmo_events_distributed (
event_id UUID,
session_uuid UUID,
event_time DateTime,
...
) ENGINE = Distributed('cluster_name', default, cmo_events, session_uuid)Distributed表的作用：- 写入：根据
session_uuid的哈希值，将数据分发到对应节点的cmo_events表 - 查询：从所有节点的
cmo_events表中拉取数据，进行聚合
- 写入：根据
四、Distributed 表的工作原理
1. 写入时自动分发
python
# 业务代码只需连接任意节点，写入 Distributed 表
client.execute("INSERT INTO cmo_events_distributed (...) VALUES", batch_data)- ClickHouse 会自动：
- 计算
session_uuid的哈希值 - 根据
cluster_name配置，决定数据写入哪个分片 - 将数据写入对应节点的本地
cmo_events表
- 计算
2. 读取时自动聚合
sql
-- 查询 Distributed 表时，ClickHouse 自动从所有分片拉取数据
SELECT COUNT(*) FROM cmo_events_distributed WHERE session_uuid = 'xxx'- ClickHouse 会：
- 向所有分片节点发送查询
- 在协调节点上聚合结果
- 返回最终结果
五、腾讯云 ClickHouse 的封装逻辑
1. 自动分片配置
- 腾讯云会自动维护
clusters.xml，你只需：- 在控制台创建集群
- 选择分片规则（如按
session_uuid哈希分片）
2. 虚拟IP 接入
- 腾讯云提供一个 统一接入点 （如
cluster.clickhouse.tencent.com） - 你只需连接这个虚拟IP，即可：
- 写入
Distributed表（自动分发） - 查询
Distributed表（自动聚合）
- 写入
3. 无需关心底层节点
- 你只需在任意节点创建
Distributed表 - 腾讯云自动维护分片节点的连接和容错
六、Distributed 表的使用示例
1. 配置集群（腾讯云自动完成）
xml
<!-- config.d/clusters.xml -->
<cluster>
<shard>
<replica><host>clickhouse-01</host></replica>
</shard>
<shard>
<replica><host>clickhouse-02</host></replica>
</shard>
</cluster>2. 创建分布式表
sql
CREATE TABLE cmo_events_distributed (
event_id UUID,
session_uuid UUID,
event_time DateTime,
...
) ENGINE = Distributed('cluster_name', default, cmo_events, session_uuid)3. 业务代码写入
python
# 连接任意节点，写入 Distributed 表
client = Client(host='cluster.clickhouse.tencent.com', port=9000, user='default', password='xxx')
client.execute("INSERT INTO cmo_events_distributed (...) VALUES", batch_data)4. 业务代码查询
python
# 查询 Distributed 表将自动聚合所有分片数据
result = client.execute("SELECT COUNT(*) FROM cmo_events_distributed WHERE session_uuid = 'xxx'")七、Distributed 表 vs Redis Cluster 的对比
| 特性 | Redis Cluster | ClickHouse Distributed 表 |
|---|---|---|
| 数据分片 | 自动（客户端透明） | 需显式创建 Distributed 表 |
| 写入透明性 | 客户端自动路由 | 需写入 Distributed 表 |
| 查询透明性 | 客户端自动路由 | 查询 Distributed 表自动聚合 |
| 容错机制 | 自动重试 | 需配合 ReplicatedMergeTree |
| 运维管理 | 完全透明 | 需配置分片规则 |
| 扩容感知 | 客户端自动感知 | 需重新配置集群（或腾讯云自动） |
| 适用场景 | 高并发键值读写 | 大规模OLAP分析 |
八、总结
-
Distributed 表的本质：
- 不存储数据
- 仅作为分片表的代理层
- 写入时自动分发，读取时自动聚合
-
使用方式：
- 写入
Distributed表（如cmo_events_distributed） - 查询
Distributed表（自动聚合所有分片）
- 写入
-
腾讯云 ClickHouse 的简化：
- 自动维护分片规则
- 提供虚拟IP接入点
- 业务代码无需关心底层节点
🚀 关键提示：
- 你只需连接到集群入口（如腾讯云提供的域名），写入
Distributed表即可实现分布式写入- 读取
Distributed表即可获取完整数据（无需手动聚合）Distributed表的_distributed后缀只是命名习惯，不是强制要求