fix: [analyzer] Local analyzer works

Pipfile

@@ -10,7 +10,7 @@ redis = "*"
 hiredis = "*"
 pyshark = "*"
 ipa = {editable = true,path = "."}
-markdown = "*"
+markdown-strings = "*"
 
 [requires]
 python_version = "3.6"

Pipfile.lock

@@ -1,7 +1,7 @@
 {
     "_meta": {
         "hash": {
-            "sha256": "5b257321cbb7c926fc129a7e5195588ed28b9a84037556db11573e3e67442438"
+            "sha256": "9313ed688a2fb9ba22b23a7dbec3e0978f1e6d1f5ff8bb7dc00651bc11b50ade"
         },
         "pipfile-spec": 6,
         "requires": {
@@ -81,10 +81,10 @@
             ],
             "version": "==4.4.1"
         },
-        "markdown": {
+        "markdown-strings": {
             "hashes": [
-                "sha256:2e50876bcdd74517e7b71f3e7a76102050edec255b3983403f1a63e7c8a41e7a",
-                "sha256:56a46ac655704b91e5b7e6326ce43d5ef72411376588afa1dd90e881b83c7e8c"
+                "sha256:141e55d48742d6c19b09500bb71760f3aa04b6532c5d7725b9a030ce90663708",
+                "sha256:21bfc2ff1e6eec1209bb5835f090f9576cec46de38fb1991f57a422bb5efcf6f"
             ],
             "index": "pypi",
             "version": "==3.1.1"

README.md

@@ -31,7 +31,7 @@ pipenv install
 
 ## Usage
 #### Start the redis server
-Don't forget to set the DB directory in the redis.conf configuration. By default, the redis for Passive DNS is running on TCP port 6400
+Don't forget to set the DB directory in the redis.conf configuration. By default, the redis for IPA is running on TCP port 6405.
 ```shell script
 ../redis/src/redis-server ./etc/redis.conf
 ```
@@ -41,8 +41,7 @@ Don't forget to set the DB directory in the redis.conf configuration. By default
 cd ./etc
 cp analyzer.conf.sample analyzer.conf
 ```
-
-Edit the analyzer.conf to match the UUID of the analyzer queue from your D4 server.
+Edit analyzer.conf to match the UUID of the analyzer queue from your D4 server.
 ```shell script
 [global]
 my-uuid = 6072e072-bfaa-4395-9bb1-cdb3b470d715
@@ -51,7 +50,7 @@ d4-server = 127.0.0.1:6380
 logging-level = INFO
 ```
 
-Then you can start the analyzer.
+#### Start the analyzer
 ```shell script
 cd ../bin
 python3 run_ipa.py

bin/export.py

@@ -0,0 +1,37 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+#
+# Markdown export module
+#
+# Copyright (C) 2019  Romain Kieffer
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU Affero General Public License as
+# published by the Free Software Foundation, either version 3 of the
+# License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU Affero General Public License for more details.
+#
+# You should have received a copy of the GNU Affero General Public License
+# along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+import markdown_strings as mds
+import redis
+import os
+
+
+analyzer_redis_host = os.getenv('D4_ANALYZER_REDIS_HOST', '127.0.0.1')
+analyzer_redis_port = int(os.getenv('D4_ANALYZER_REDIS_PORT', 6405))
+r = redis.Redis(host=analyzer_redis_host, port=analyzer_redis_port)
+
+
+def export_icmp_types():
+    res = mds.table_row(['ICMP Type', 'Count'], [10, 10]) + '\n'
+    res += '| :----- | -----: |\n'
+    redis_dict = r.hgetall('icmp')
+    for key in redis_dict:
+        res += mds.table_row([key.decode(), redis_dict[key].decode()], [10, 10]) + '\n'
+    return res

bin/run_ipa.py

@@ -26,7 +26,7 @@ import argparse
 if __name__ == "__main__":
 
     parser = argparse.ArgumentParser(description='D4-IPA')
-    parser.add_argument('-p', '--path', type=int, nargs=1, help='Path of local dataset.')
+    parser.add_argument('-p', '--path', type=str, nargs=1, help='Path of local dataset.')
 
     dataset = None
 

etc/redis.conf

@@ -89,7 +89,7 @@ protected-mode yes
 
 # Accept connections on the specified port, default is 6379 (IANA #815344).
 # If port 0 is specified Redis will not listen on a TCP socket.
-port 6400
+port 6405
 
 # TCP listen() backlog.
 #

lib/analyzer.py

@@ -25,7 +25,7 @@ import configparser
 import logging
 
 from lib.inspection import get_cap, get_protocol, check_icmp_checksum, get_icmp_payload, get_icmp_ip, \
-    unassigned_icmp_types, deprecated_icmp_types, get_src_port, get_dst_port, list_caps
+    unassigned_icmp_types, deprecated_icmp_types, get_src_port, get_dst_port, list_caps, init_cap_list
 
 
 class Analyzer:
@@ -51,7 +51,7 @@ class Analyzer:
         self.logger.addHandler(self.ch)
 
         analyzer_redis_host = os.getenv('D4_ANALYZER_REDIS_HOST', '127.0.0.1')
-        analyzer_redis_port = int(os.getenv('D4_ANALYZER_REDIS_PORT', 6400))
+        analyzer_redis_port = int(os.getenv('D4_ANALYZER_REDIS_PORT', 6405))
         self.r = redis.Redis(host=analyzer_redis_host, port=analyzer_redis_port)
 
         self.dataset = dataset_path
@@ -65,9 +65,15 @@ class Analyzer:
             self.r_d4 = redis.Redis(host=host_redis_metadata, port=port_redis_metadata, db=2)
         else:
             self.logger.info("Starting local analyzer")
-            self.update_queue()
+            self.queue = "to_scan"
             self.cap_list = []
+            self.logger.info("Adding dataset caps to local queue")
+            self.cap_list = init_cap_list(self.dataset)
+            self.logger.info(len(self.cap_list))
+            self.update_queue()
+            self.logger.info("Processing...")
             self.process_local()
+            self.logger.info("Done.")
             time.sleep(15)
             c = self.update_queue()
             if c == 0:
@@ -83,27 +89,23 @@ class Analyzer:
         p.execute()
 
     def update_queue(self):
-        """
-        Each parser instance is given a list of days, and thus a list of caps to parse.
-        This method lets the parser confront his list of caps with the caps in his queue.
-        """
         remaining_caps = list_caps(self.queue, self.r)
         current_caps = list_caps('scanning', self.r)
         parsed_caps = list_caps('scanned', self.r)
         caps_to_add = []
         if remaining_caps:
-            print('[*] Queue already populated.')
+            self.logger.info('Queue already populated.')
             if self.cap_list:
                 for cap in self.cap_list:
                     if cap not in remaining_caps and cap not in parsed_caps and cap not in current_caps:
                         caps_to_add.append(cap)
             if not caps_to_add:
-                print('[*] Already up to date.')
+                self.logger.info('Already up to date.')
                 return 1
-            print('[o] Queue updated.')
+            self.logger.info('Queue updated.')
         else:
             if self.cap_list:
-                print('[*] No caps, initializing...')
+                self.logger.info('No caps enqueued, initializing...')
                 caps_to_add = self.cap_list
             elif current_caps:
                 return 0
@@ -115,10 +117,10 @@ class Analyzer:
         Dissects the cap file to extract info.
         """
         if cap is None:
-            print('[X] No caps to parse!')
+            self.logger.info('[X] No caps to parse!')
             return 0
 
-        print('[*] Started parsing...')
+        self.logger.info('Parsing cap ' + cap.input_filename[-15:])
 
         pipeline = self.r.pipeline()
         for packet in cap:
@@ -167,9 +169,9 @@ class Analyzer:
 
     def pop_cap(self):
         if not self.dataset:
-            absolute_path = self.r_d4.rpop(self.queue)
+            absolute_path = self.r_d4.rpop(self.queue).decode()
         else:
-            absolute_path = self.r.rpop('to_scan')
+            absolute_path = self.r.lpop('to_scan').decode()
         return get_cap(absolute_path)
 
     def process_d4(self):

lib/inspection.py

@@ -129,16 +129,15 @@ def get_files(path) -> list:
     return caps
 
 
-def init_cap_list(dataset_path: str, daylist: list) -> list:
-    cap_list = []
-    if not daylist:
-        return []
-    for day in daylist:
-        cap_path = dataset_path + str(day) + '/*.gz'
+def init_cap_list(dataset_path: str) -> list:
+    if dataset_path[-1] == '/':
+        extension = '*.gz'
+    else:
+        extension = '/*.gz'
+    cap_path = dataset_path + extension
     caps = get_files(cap_path)
     caps.sort()
-        cap_list += caps
-    return cap_list
+    return caps
 
 
 def list_caps(state: str, redis):

setup.py

@@ -12,7 +12,7 @@ setup(
     url='https://github.com/D4-project/analyzer-d4-ipa',
     description='Pcap icmp parser focused on DDoS detection',
     packages=['lib'],
-    scripts=[],
+    scripts=['bin/run_ipa.py', 'bin/export.py'],
     include_package_data=True,
     classifiers=[
         'License :: OSI Approved :: BSD License',