python連接clickhouse數據庫的兩種方式小結

　　python連接clickhouse數據庫

　　在python中獲取系統信息的一個好辦法是使用psutil這個第三方模塊。

　　顧名思義，psutil = process and system utilities，它不僅可以通過一兩行代碼實現系統監控，還可以跨平臺使用。

　　主要針對clickhouse_driver的使用進行簡要介紹

第一步：

　　通過pip install clickhouse_driver 安裝 clickhouse_driver

第二步：

　　方法一：使用clickhouse_driver 包中的client類,通過實例化一個客戶端進行對數據庫的增刪改查操作

from clickhouse_driver import client
from datetime import datetime
import psutil
host_name = '192.168.50.94'
client = client(host=host_name,database='default',user='default',password='自己設的密碼',send_receive_timeout=20,port=55666)
now = datetime.now()
time_stamp = now.strftime('%a %b %d %h:%m:%s cst %y')# tue apr 06 15:32:55 cst 2021 <class 'str'>
create_at = datetime.now().strftime('%y-%m-%d %h:%m:%s')
disk_io = psutil.disk_io_counters()
net_io = psutil.net_io_counters()
chart_name = ["磁盤io","網絡io"]
metric_name1 = ["讀(數量)","寫(數量)", "讀(字節)", "寫(字節)", "讀(時間)", "寫(時間)"]
metric_name2 = ["發送字節數","接收字節數","發送包數","接收包"]
metric_value1 = [disk_io.read_count,disk_io.write_count,disk_io.read_bytes,disk_io.write_bytes,disk_io.read_time,disk_io.write_time]
metric_value2 = [net_io.bytes_sent,net_io.bytes_recv,net_io.packets_sent,net_io.packets_recv]
try:
for i in chart_name:
if i is "磁盤io":
for j in metric_name1:
sql = "insert into clickhouse_host_metrics777(time_stamp,host_name, chart_name, metric_name,metric_value,create_at) " \
"values('%s','%s','%s','%s','%s','%s')" % \
(time_stamp, host_name, i, j, metric_value1[metric_name1.index(j)], create_at)
res = client.execute(sql)
elif i is "網絡io":
for j in metric_name2:
sql = "insert into clickhouse_host_metrics777(time_stamp,host_name, chart_name, metric_name,metric_value,create_at) " \
"values('%s','%s','%s','%s','%s','%s')" % \
(time_stamp, host_name, i, j, metric_value2[metric_name2.index(j)], create_at)
res = client.execute(sql)
print("成功寫入數據")
except exception as e:
print(str(e))

　　方法二：使用clickhouse_driver 包中的connect函數,通過實例化一個客戶端進行對數據庫的增刪改查操作

from datetime import datetime
import psutil
from clickhouse_driver import connect
host_name = '192.168.50.94'
#賬號:密碼@主機名:端口號/數據庫
conn = connect('clickhouse://default:自己設的密碼@'+host_name+':55666/default')
cursor = conn.cursor()
now = datetime.now()
time_stamp = now.strftime('%a %b %d %h:%m:%s cst %y')# tue apr 06 15:32:55 cst 2021 <class 'str'>
create_at = datetime.now().strftime('%y-%m-%d %h:%m:%s')
disk_io = psutil.disk_io_counters()
net_io = psutil.net_io_counters()
chart_name = ["磁盤io","網絡io"]
metric_name1 = ["讀(數量)","寫(數量)", "讀(字節)", "寫(字節)", "讀(時間)", "寫(時間)"]
metric_name2 = ["發送字節數","接收字節數","發送包數","接收包"]
metric_value1 = [disk_io.read_count,disk_io.write_count,disk_io.read_bytes,disk_io.write_bytes,disk_io.read_time,disk_io.write_time]
metric_value2 = [net_io.bytes_sent,net_io.bytes_recv,net_io.packets_sent,net_io.packets_recv]
try:
for i in chart_name:
if i is "磁盤io":
for j in metric_name1:
sql = "insert into clickhouse_host_metrics777(time_stamp,host_name, chart_name, metric_name,metric_value,create_at) values('%s','%s','%s','%s','%s','%s')" % \
(time_stamp, host_name, i, j, metric_value1[metric_name1.index(j)], create_at)
# res = client.execute(sql)
res = cursor.execute(sql)
elif i is "網絡io":
for j in metric_name2:
sql = "insert into clickhouse_host_metrics777(time_stamp,host_name, chart_name, metric_name,metric_value,create_at) values('%s','%s','%s','%s','%s','%s')" % \
(time_stamp, host_name, i, j, metric_value2[metric_name2.index(j)], create_at)
res = cursor.execute(sql)
cursor.close()
print("成功寫入數據")
except exception as e:
print(str(e))

　　python將數據寫入clickhouse
 

from clickhouse_driver import client

# connect clickhouse
client = client(host= ,port= ,user= ,database= , password=)


# 得到table1中查詢的數據導入table2中（database2中應該事先建立對應的table2表）
query_ck_sql = """ select *
from database1.table1
where date = today() """
# 導入數據到臨時表
try:
# 導入數據
client.execute("insert into {official_table_db}.{official_all_table_name} \
{query_ck_sql}".format(
official_table_db = database2,
official_table_name = table2,
query_ck_sql = query_ck_sql)
,types_check = true)
except exception as e:
print str(e)

　　以上為個人經驗，希望能給大家一個參考，也希望大家多多支持