diff --git a/README.md b/README.md
index dd20096..cb3b994 100644
--- a/README.md
+++ b/README.md
@@ -18,15 +18,15 @@ We include all known proxies of listed sites to raise chance of correct routing.
 - facebook.com - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 - instagram.com - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 - meta - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
-- netflix - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
-- openai.com
-- x.com - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
-- youtube.com - from [[x]](https://github.com/touhidurrr/iplist-youtube?tab=readme-ov-file) [[x]](https://www.gstatic.com/ipranges/goog.json) [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 - microsoft - from [[x]](https://networksdb.io/ip-addresses-of/microsoft-corp)
+- netflix - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 - nhentai.net - from [[x]](https://networksdb.io/domain-to-ips/nhentai.net)
+- openai.com
 - rutracker.org - from [[x]](https://networksdb.io/)
+- x.com - from [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 - xhamster.com - from [[x]](https://networksdb.io/)
 - xvideos.com - from [[x]](https://networksdb.io/)
+- youtube.com - from [[x]](https://github.com/touhidurrr/iplist-youtube?tab=readme-ov-file) [[x]](https://www.gstatic.com/ipranges/goog.json) [[x]](https://rockblack.su/vpn/dopolnitelno/diapazon-ip-adresov)
 <div id="auto-sort-end"/>
 
 Not used DBs, but ready to deploy:
diff --git a/src/sort_db.py b/src/sort_db.py
index 4b182b4..5e179df 100644
--- a/src/sort_db.py
+++ b/src/sort_db.py
@@ -3,6 +3,7 @@
 import os
 from os import listdir
 from os.path import isfile, join
+import pandas as pd
 
 
 def drop_duplicates(data):
@@ -14,11 +15,9 @@ def drop_duplicates(data):
     not_cidr_data = data[~data['ipv4'].str.contains('/')]
     for index, row in cidr_data.iterrows():
         try:
-            ip = ipaddress.ip_network(row['ipv4'], strict=False)
-            for i in ip.hosts():
-                if str(i) in not_cidr_data['ipv4'].values:
-                    not_cidr_data = not_cidr_data[~(not_cidr_data['ipv4'].astype(str) == str(i))]
-                    log_info(f'Dropped {str(i)} because it is included in {row["ipv4"]}')
+            ip_net = ipaddress.ip_network(row['ipv4'], strict=False)
+            not_cidr_data.loc[:, 'ipv4'] = not_cidr_data['ipv4'].apply(lambda x: None if ipaddress.ip_address(x) in ip_net else x)
+            not_cidr_data = not_cidr_data.dropna(subset=['ipv4'])
         except ValueError as e:
             log_warning(f"Invalid CIDR notation {row['ipv4']}: {e}")
             # remove invalid CIDR
@@ -66,25 +65,17 @@ def drop_duplicates_in_known(data):
     if len(not_cidr_data) == 0:
         log_info("No non-CIDR data found")
     else:
-        cidr_ips = set()
         for cidr in cidr_data.iloc[:, 0]:
             try:
                 ip_network = ipaddress.ip_network(cidr, strict=False)
-                cidr_ips.update(str(ip) for ip in ip_network.hosts())
+                not_cidr_data['ipv4'] = not_cidr_data.iloc[:, 0].apply(lambda x: None if ipaddress.ip_address(x) in ip_network else x)
+                not_cidr_data = not_cidr_data.dropna(subset=[not_cidr_data.columns[0]])
             except ValueError as e:
                 log_warning(f"Invalid CIDR notation {cidr}: {e}")
                 # remove invalid CIDR
                 cidr_data = cidr_data[~(cidr_data.iloc[:, 0].astype(str) == cidr)]
                 log_info(f'Dropped {cidr} because it is invalid CIDR notation')
 
-
-        not_cidr_data = not_cidr_data[~not_cidr_data.iloc[:, 0].astype(str).isin(cidr_ips)]
-
-        dropped_ips = set(original_data.iloc[:, 0].astype(str)) - set(not_cidr_data.iloc[:, 0].astype(str)) - set(cidr_data.iloc[:, 0].astype(str))
-        for ip in dropped_ips:
-            log_info(f'Dropped {ip} because it is included in a CIDR range')
-
-
     data = pd.concat([not_cidr_data, cidr_data], ignore_index=True)
     if len(original_data) != len(data):
         log_happy(f"Dropped {len(original_data) - len(data)} duplicate IP addresses")