ping 3000 servers with 10 timeout in 150 sec

async_net_scanner.py

@@ -1,7 +1,7 @@
 import asyncio
 import time
 from socket import gaierror
-from typing import List, Tuple, Union
+from typing import List, Tuple, Union, Dict, Optional
 import aiohttp
 from aiohttp.client_exceptions import TooManyRedirects
 import logging
@@ -9,10 +9,18 @@ import sys
 from collections import Counter
 from functools import wraps
 
+HEADERS = {
+    'user-agent': 'Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:91.0) Gecko/20100101 Firefox/91.0 ', }
 
 console_logger = logging.getLogger(__name__)
+formatter = logging.Formatter(datefmt="%Y.%m.%d %H:%M:%S",
+                              fmt='%(asctime)s | message: %(message)s')
+                              # fmt='%(asctime)s | %(levelname)s | process: %(process)d | module name: %(name)s | '
+                              #     'func name: %(funcName)s | line number: %(lineno)s | message: %(message)s',)
+handler = logging.StreamHandler(sys.stdout)
+handler.setFormatter(formatter)
 console_logger.setLevel(logging.INFO)
-console_logger.addHandler(logging.StreamHandler(sys.stdout))
+console_logger.addHandler(handler)
 
 
 def time_to_execute(time_form='sec'):
@@ -34,19 +42,22 @@ def time_to_execute(time_form='sec'):
             exec_time = (end - begin) / multiply[time_form]
             console_logger.info(f'Duration {func.__name__}, {time_form}: {exec_time}')
             return result
+
         return new_func
+
     return wrapper
 
+
 class NetScanner:
     """
     Scan urls asyncable
-
     :param logger: Setup logger | logging.Logger
     """
 
-    def __init__(self, logger: logging.Logger) -> None:
+    def __init__(self, sem: int, logger: logging.Logger) -> None:
         self.logger = logger
-        self.sem = asyncio.Semaphore(50)
+        self.sem = asyncio.Semaphore(sem)
+        self.count = 0
 
     async def _head_request(self, session: aiohttp.ClientSession, url: str, timeout: int) -> Union[str, None]:
         """
@@ -54,12 +65,13 @@ class NetScanner:
         :param session: Http session for request | aiohttp.ClientSession
         :param ip: Get ip which server will be checked | str
         :param timeout: set timeout to wait in seconds | int
-
         :return: Ip with status OK | str
         """
         try:
             # A HEAD request is quicker than a GET request
             async with self.sem:
+                self.count += 1
+                console_logger.info(f'Count of total requests: {self.count}')
                 resp = await session.head(url, allow_redirects=True, ssl=False, timeout=timeout)
                 async with resp:
                     status = resp.status
@@ -96,23 +108,22 @@ class NetScanner:
         Get list of ip addresses and return which of them is available
         :param ip_addresses: list of ip addresses | List[str]
         :param timeout: set timeout to wait in seconds | int
-
         :return: List of ip addresses | List[str]
         """
         # Next setting is configure dns and limit of connections for session like
-        # aiohttp.ClientSession(connector=connector)
-        # connector = aiohttp.TCPConnector(limit=1000, ttl_dns_cache=300)
+        # aiohttp.ClientSession(connector=connector, shared_timeout=timeout)
+        connector = aiohttp.TCPConnector(limit=500)
+        # shared_timeout = aiohttp.ClientTimeout(total=5)
 
-        async with aiohttp.ClientSession() as session:
+        async with aiohttp.ClientSession(connector=connector) as session:
             servers = await asyncio.gather(*[self._head_request(session, ip, timeout) for ip in ip_addresses])
             return list(servers)
 
-    def get_potential_servers_to_discover(self, ip_addresses: List[str], timeout) -> Tuple[str]:
+    def ping_servers(self, ip_addresses: List[str], timeout) -> Tuple[str]:
         """
         Loops through the ip addresses, head request each, append available to potential servers tuple
         :param ip_addresses: bunch of ip addresses | list []
         :param timeout: set timeout to wait in seconds | int
-
         :return: available servers | tuple()
         """
         self.logger.info("Started polling potential servers")
@@ -126,8 +137,7 @@ class NetScanner:
         return tuple(servers)
 
 
-server_checker = NetScanner(console_logger)
-urls = ['https://httpbin.org/delay/3' for _ in range(4000)]
-available_urls = server_checker.get_potential_servers_to_discover(urls, timeout=4)
-
-print(f'{Counter(available_urls).most_common(2)}')
+server_checker = NetScanner(200, console_logger)
+urls = ['https://httpbin.org/delay/10' for _ in range(3000)]
+available_urls = server_checker.ping_servers(urls, timeout=15)
+console_logger.info(f'Ping result: {Counter(available_urls).most_common(2)}')