主要涉及两个sql
sql1：
select t1.hostid,t1.name from hosts t1 join hosts_groups t2 on t1.hostid=t2.hostid where t1.available=1 and t1.status=0 and t2.groupid in(1,2);
t2.groupid in(1,2)为主机群组的groupid，这里举例groupid为1和2，可以直接查询groups表获取groupid。如果不区分群组可以去掉这个条件。
sql2：
SELECt t3.name,t1.value FROM history t1 join items t2 on t1.itemid=t2.itemid join hosts t3 on t3.hostid=t2.hostid WHERe t2.key_="vfs.fs.size[/export,pfree]" and t3.hostid=%s order by t1.clock DESC LIMIT 1;
t2.key_是监控项的表达式，这里是vfs.fs.size[/data,pfree]，也就是/data路径下的剩余空间的占比，更改为自己想要的路径即可，也可以使用其他监控项，比如内存cpu等等。t3.hostid的值是sql1查询出的hostid，循环查询即可。
以下是根据这两个sql写的python脚本，可以直接得到结果
import pymysql
def execute():
disk_info = []
sql1 = """
select t1.hostid,t1.name from hosts t1 join hosts_groups t2 on
t1.hostid=t2.hostid where t1.available=1 and t1.status=0 and t2.groupid in(1,1);
"""
db = pymysql.connect(host='10.10.10.10', user='root', password='mypassword',
database='zabbix',
port=3306, charset='utf8')
cursor1 = db.cursor()
cursor1.execute(sql1)
host_info = cursor1.fetchall()
for i in host_info:
sql2 = """
SELECt t3.name,t1.value FROM history t1 join items t2 on t1.itemid=t2.itemid
join hosts t3 on t3.hostid=t2.hostid WHERe t2.key_="vfs.fs.size[/export,pfree]"
and t3.hostid=%s order by t1.clock DESC LIMIT 1;
""" % i[0]
cursor2 = db.cursor()
cursor2.execute(sql2)
disk_info2 = cursor2.fetchall()
disk_info.append(disk_info2)
db.close()
for j in disk_info:
try:
print(j[0])
except:
print(j)
execute()
