tcpsnitch_analyzer 0.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 50b9dac312c23a4ec6e10f497b675078c0d6e7cf
+  data.tar.gz: 86e2119b21841973b78cadca2965d89377500ee4
+SHA512:
+  metadata.gz: b8446c031fe67e81e5cbd72ab6b0bd3dcd2bc827d023684bc2f3b59da73e6e7566d6d415938d417cd71fdf18319940ed0236a526988291134bb063aa811d5e8b
+  data.tar.gz: 58b7ef0d840080ad862399c4a55ec8b1b01d15ad3c33fe7ef94e0aeaf916618a299cc03d590f6234b2ed7244e7f9417fb5baa16b7311566d45b745e30bbac974

data/bin/tcpsnitch_analyzer

@@ -0,0 +1,24 @@
+#!/usr/bin/env ruby
+
+require 'tcpsnitch_analyzer'
+
+# Parse options
+options = TcpsnitchAnalyzer::OptParser.parse(ARGV)
+
+# Check files
+files = ARGV
+
+if files.empty? then 
+  TcpsnitchAnalyzer::error("missing argument, must provide at least 1 file")
+  exit 1
+end
+
+if options.analysis_type == TcpsnitchAnalyzer::TimeSerieStat and 
+   files.size > 1 then
+  TcpsnitchAnalyzer::error("invalid argument: time serie analysis has no "\
+                            "meaning on multiple JSON traces")
+  exit 1
+end
+
+# Start processing
+TcpsnitchAnalyzer.process_files(options, files)

data/lib/tcpsnitch_analyzer.rb

@@ -0,0 +1,101 @@
+require 'oj'
+
+require 'tcpsnitch_analyzer/opt_parser'
+require 'tcpsnitch_analyzer/descriptive_stat'
+require 'tcpsnitch_analyzer/proportion_stat'
+require 'tcpsnitch_analyzer/time_serie_stat'
+
+module TcpsnitchAnalyzer
+  EXECUTABLE = 'tcpsnitch_analyzer'
+  VERSION = '0.0.0'
+  
+  # Configure Oj
+  Oj.default_options = { symbol_keys: true }
+
+  class << self
+    def process_files(options, files)
+      # We DO NOT want to read the entire JSON files into memory.
+      # We DO NOT want to build a Ruby object for the entire JSON array.
+      #
+      # Instead, we want to the file line by line, where each line consists of 
+      # a single event. We then instantiate each event individually and discard 
+      # them as we consume the file, thus giving O(1) memory consumption.
+      matched_data = false
+
+      files.each do |file|
+        # IO.each should not read entire file in memory. To verify?
+        File.open(file).each_with_index do |line, index|
+          next if index == 0    # First line is opening '[' of JSON Array
+          next if line.eql? "]" # Last line is closing ']' of JSON Array
+
+          # Parse JSON object
+          begin 
+            hash = Oj.load(line.chomp(",\n")) # Remove ',\n' after JSON object
+          rescue Exception => e 
+            error(e)
+          end
+         
+          # Skip if filter does not match
+          next unless filter(hash, options.event_filter)  
+          matched_data = true
+          
+          # Extract value
+          begin
+            val = node_val(hash, options.node_path)
+          rescue Exception => e
+            error("invalid -n argument: '#{options.node_path}'")
+          end
+
+          # Compute on value
+          if options.analysis_type == TimeSerieStat
+            options.analysis_type.add_point(node_val(hash, 'timestamp'), val)
+          else
+            options.analysis_type.add_val(val)
+          end
+        end
+      end
+
+      # Output results
+      puts_options_header(options)
+      if matched_data 
+        options.analysis_type.print(options)
+      else
+        puts "No data point matching criterias."
+      end
+    end
+    
+    def error(msg)
+      puts "#{EXECUTABLE}: #{msg}."
+      puts "Try '#{EXECUTABLE} -h' for more information."
+      exit 1
+    end
+
+    def filter(hash, filter)
+      if filter then
+        hash[:type] == filter ? hash : nil
+      else
+        hash
+      end
+    end
+
+    def val_for(hash, keys)
+        keys.reduce(hash) { |h, key| h[key] }
+    end
+
+    def keys_from_path(path)
+      path.split('.').collect(&:to_sym)
+    end
+
+    def node_val(hash, path)
+      val_for(hash, keys_from_path(path))
+    end
+
+    def puts_options_header(options)
+      puts "JSON node:".ljust(15) + "#{options.node_path}"
+      event_filter = options.event_filter ? options.event_filter : "/" 
+      puts "Type filter:".ljust(15) + event_filter
+      puts ""
+    end
+
+  end # class << self
+end # module

data/lib/tcpsnitch_analyzer/descriptive_stat.rb

@@ -0,0 +1,53 @@
+require 'descriptive_statistics'
+require 'gnuplot'
+
+module TcpsnitchAnalyzer
+  class DescriptiveStat
+    @@data = []
+
+    def self.add_val(val)
+      if val.is_a? Integer
+        @@data.push(val)
+      else
+        if val.is_a? Hash
+          TcpsnitchAnalyzer.error("invalid value for descriptive statistic: "\
+                                  "non-terminal node")
+        else
+          TcpsnitchAnalyzer.error("invalid value for descriptive statistic: '#{val}'")
+        end
+      end
+    end
+
+    def self.print(options)
+      puts "Descriptive statistics:"
+      @@data.descriptive_statistics.each do |key, value|
+        puts "#{key}".ljust(20) + "#{value}" 
+      end
+      
+      # Only plot CDF is we have a range
+      return unless @@data.range > 0 
+
+      x = @@data.sort
+      n = x.size
+      y = x.map do |el|
+        ((x.rindex { |v| v <= el } || -1.0) + 1.0) / n * 100.0
+      end
+
+      Gnuplot.open do |gp|
+        Gnuplot::Plot.new(gp) do |plot|
+          plot.xrange "[#{@@data.min}:#{@@data.max}]; set logscale x"
+          plot.title  "CDF for #{options.node_path} (#{options.event_filter} events)"
+          plot.xlabel "Value"
+          plot.ylabel "Normal CDF"
+          plot.data << Gnuplot::DataSet.new([x,y]) do |ds|
+            ds.with = "lines"
+            ds.linewidth = 4
+            ds.title = options.node_path.split('.').last
+          end
+        end
+      end # Gnuplot.open
+    end # def self.print
+
+  end # class
+end # module
+

data/lib/tcpsnitch_analyzer/opt_parser.rb

@@ -0,0 +1,65 @@
+require 'optparse'
+require 'ostruct'
+
+module TcpsnitchAnalyzer
+  class OptParser 
+    def self.parse(args)
+      options = OpenStruct.new
+      options.analysis_type = ProportionStat
+      options.event_filter = nil 
+      options.node_path = "type"
+        
+      begin
+        OptionParser.new do |opts|
+          opts.banner = "Usage: #{EXECUTABLE} [-h] [options] file...\n"
+          opts.separator ""
+          opts.separator "Analyze tcpsnitch JSON traces."
+          opts.separator ""
+          opts.separator "Options:"
+          
+          opts.on("-a", "--analysis [TYPE]", "TYPE of statistic analysis: 
+                  descriptive (d), proportion (p) or timeserie (t).") do |type|
+            TcpsnitchAnalyzer.error("missing -a argument") unless type 
+
+            case type.downcase
+            when /^d.*/
+              options.analysis_type = DescriptiveStat
+            when /^p.*/
+              options.analysis_type = ProportionStat
+            when /^t.*/
+              options.analysis_type = TimeSerieStat
+            else
+              TcpsnitchAnalyzer.error("invalid -a argument: '#{type}'")
+            end
+          end
+   
+          opts.on("-e", "--event [FILTER]", "filter on events of type FILTER") do |ev|
+            TcpsnitchAnalyzer.error("missing -e argument") unless ev 
+            options.event_filter = ev          
+          end
+
+          opts.on_tail("-h", "--help", "show this help text") do 
+            puts opts
+            exit
+          end
+   
+          opts.on("-n", "--node [PATH]", "compute on node at PATH") do |node| 
+            TcpsnitchAnalyzer.error("missing -n argument") unless node 
+            options.node_path = node 
+          end
+   
+          opts.on_tail("--version", "show version") do 
+            puts VERSION
+            exit
+          end
+
+        end.parse!(args) # OptionParser
+      rescue OptionParser::ParseError => e 
+        TcpsnitchAnalyzer.error(e)
+      end
+
+      options
+    end # def self.parse
+
+  end # class
+end # module

data/lib/tcpsnitch_analyzer/proportion_stat.rb

@@ -0,0 +1,22 @@
+module TcpsnitchAnalyzer
+  class ProportionStat
+    @@count = 0
+    @@hash = Hash.new(0)
+
+    def self.add_val(val)
+      @@count += 1
+      @@hash[val] += 1
+    end
+
+    def self.print(options)
+      puts "Proportion statistics:"
+
+      @@hash.sort_by { |val, count| -count }.each do |val, count|
+        pc = ((count.to_f/@@count) * 100).round(2)
+        puts "#{pc}%".ljust(7) + "(#{count})".ljust(10) + "#{val}"
+      end
+      puts "100%".ljust(7) + "(#{@@count})" 
+    end
+
+  end # class
+end # module

data/lib/tcpsnitch_analyzer/time_serie_stat.rb

@@ -0,0 +1,32 @@
+module TcpsnitchAnalyzer
+  class TimeSerieStat
+    @@x = []
+    @@y = []
+
+    def self.add_point(timestamp, y)
+      usec = timestamp[:sec] * 1000000 + timestamp[:usec]
+      @@min ||= usec
+      @@x.push(usec-@@min)
+      @@y.push(y)
+    end
+
+    def self.print(options)
+      puts "Time serie plot"
+      
+      Gnuplot.open do |gp|
+        Gnuplot::Plot.new(gp) do |plot|
+          plot.title  "Time serie: #{options.node_path}(t)"
+          plot.xlabel "Micro seconds"
+          plot.ylabel options.node_path.split('.').last.capitalize
+        
+          plot.data << Gnuplot::DataSet.new([@@x,@@y]) do |ds|
+            ds.with = "lines"
+            ds.linewidth = 4
+            ds.notitle
+          end
+        end
+      end # Gnuplot.open
+    end # def self.print
+
+  end # class
+end # module

metadata

@@ -0,0 +1,92 @@
+--- !ruby/object:Gem::Specification
+name: tcpsnitch_analyzer
+version: !ruby/object:Gem::Version
+  version: 0.0.0
+platform: ruby
+authors:
+- Gregory Vander Schueren
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2017-01-10 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: oj
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.18'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.18'
+- !ruby/object:Gem::Dependency
+  name: descriptive_statistics
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.5'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.5'
+- !ruby/object:Gem::Dependency
+  name: gnuplot
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.6'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.6'
+description: Tool to analyze tcpsnitch traces. Longer description to come.
+email: gregory.vanderschueren@gmail.com
+executables:
+- tcpsnitch_analyzer
+extensions: []
+extra_rdoc_files: []
+files:
+- bin/tcpsnitch_analyzer
+- lib/tcpsnitch_analyzer.rb
+- lib/tcpsnitch_analyzer/descriptive_stat.rb
+- lib/tcpsnitch_analyzer/opt_parser.rb
+- lib/tcpsnitch_analyzer/proportion_stat.rb
+- lib/tcpsnitch_analyzer/time_serie_stat.rb
+homepage: http://rubygems.org/gems/tcpsnitch_analyzer
+licenses:
+- GPL-3.0
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.6.8
+signing_key: 
+specification_version: 4
+summary: Tool to analyze tcpsnitch traces
+test_files: []