flapjack 0.7.22 → 0.7.25

data/CHANGELOG.md

@@ -1,5 +1,24 @@
 ## Flapjack Changelog
 
+# 0.7.25 - 2013-09-13
+- Bug: EntityCheck last_update= isn't being called for update_state since refactoring gh-303 (@ali-graham)
+- Bug: flapjack-nagios-receiver is double-escaping its JSON data gh-304 (@jessereynolds)
+
+# 0.7.24 - 2013-09-12
+- Bug: gem install of flapjack 0.7.23 fails with tzinfo-data dependency error gh-302 (@jessereynolds)
+
+# 0.7.23 - 2013-09-12
+- Bug: Quick ok -> warning -> ok -> warning triggers too many recovery notifications gh-119 (@jessereynolds)
+- Bug: Blackhole notification rule doesn't block recovery notifications gh-282 (@jessereynolds)
+- Chore: Shorten SMS messages to 159 chars on the Messagenet gateway gh-278 (@ali-graham)
+- Chore: flapjack-nagios-receiver should use Event#add gh-275 (@ali-graham)
+- Chore: Non-zero exit code after receiving SIGINT gh-266 (@ali-graham)
+- Bug: Email notifications - remove "(about a minute ago)" and fix previous state fields gh-258 (@ali-graham)
+- Chore: refactor delays filter, remove mass client failures filter gh-293 (@jessereynolds)
+- Bug: creation of scheduled maintenance fails from web UI gh-296 (@ali-graham)
+- Feature: flapjack UI needs a favicon gh-297 (@jessereynolds)
+- Chore: email notification styling gh-298 (@jessereynolds)
+
 # 0.7.22 - 2013-08-08
 - Bug: fix potential exception in json serialisation of tags in notifications gh-281 (@jessereynolds)
 

data/bin/flapjack

@@ -146,10 +146,12 @@ when "start"
     puts "Flapjack is already running."
   else
     print "Flapjack starting..."
+    return_value = nil
     runner.execute(:daemonize => daemonize) {
-      flapjack_coord.call
+      return_value = flapjack_coord.call
     }
     puts " done."
+    exit(return_value + 128) unless (return_value.nil? || (return_value == 0))
   end
 
 when "stop"

data/bin/flapjack-nagios-receiver

@@ -14,6 +14,7 @@ Oj.default_options = { :indent => 0, :mode => :strict }
 require 'dante'
 
 require 'flapjack/configuration'
+require 'flapjack/data/event'
 
 def pike(message)
   puts "piking out: #{message}"
@@ -47,16 +48,16 @@ def process_input(opts)
       state = 'ok'       if state.downcase == 'up'
       state = 'critical' if state.downcase == 'down'
       details = check_long_output ? check_long_output.gsub(/\\n/, "\n") : nil
-      event = Oj.dump({
+      event = {
         'entity'    => entity,
         'check'     => check,
         'type'      => 'service',
         'state'     => state,
         'summary'   => check_output,
         'details'   => details,
-        'timestamp' => timestamp,
-      })
-      redis.lpush 'events', event
+        'time'      => timestamp,
+      }
+      Flapjack::Data::Event.add(event, :redis => redis)
     end
   rescue Redis::CannotConnectError
     puts "Error, unable to to connect to the redis server (#{$!})"

data/bin/receive-events

@@ -12,6 +12,7 @@ require 'oj'
 Oj.default_options = { :indent => 0, :mode => :strict }
 
 require 'flapjack/configuration'
+require 'flapjack/data/event'
 
 def pike(message)
   puts "piking out: #{message}"
@@ -19,8 +20,7 @@ def pike(message)
 end
 
 def send_event(event, opts)
-  redis = opts[:redis]
-  redis.lpush 'events', event
+  Flapjack::Data::Event.add(event, :redis => opts[:redis])
 end
 
 def receive(opts)

data/features/events.feature

@@ -123,7 +123,12 @@ Feature: events
 
   @time
   Scenario: Check critical to ok when acknowledged
-    Given the check is in a critical state
+    Given the check is in an ok state
+    When  a critical event is received
+    And   one minute passes
+    And   a critical event is received
+    Then  a notification should be generated
+    # the above all needs to be just a call to the "Check ok to critical for 1 minute" Scenario if that's possible
     When  an acknowledgement event is received
     Then  a notification should be generated
     When  1 minute passes
@@ -297,101 +302,102 @@ Feature: events
 
 # commenting out this test for now, will revive it
 # when working on gh-119
-# @time
-# Scenario: a lot of quick ok -> warning -> ok -> warning
-#     Given the check is in an ok state
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   a warning event is received
-#     Then  a notification should not be generated
-#     When  10 seconds passes
-#     And   an ok event is received
-#     Then  a notification should not be generated
+@time
+Scenario: a lot of quick ok -> warning -> ok -> warning
+     Given the check is in an ok state
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  20 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should be generated
+     # recovered
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   a warning event is received
+     Then  a notification should not be generated
+     When  10 seconds passes
+     And   an ok event is received
+     Then  a notification should not be generated
 
   Scenario: scheduled maintenance created for initial check reference
     Given the check has no state
     When  an ok event is received
-    Then  scheduled maintenance should be generated
+    Then  scheduled maintenance should be generated

data/features/notification_rules.feature

@@ -45,6 +45,7 @@ Feature: Notification rules on a per contact basis
       |          |      | email         | email            |                   |                    |                   |
       |          |      | sms           | sms              |                   |                    |                   |
       | bar      |      | email         | email,sms        |                   |                    |                   |
+      | bar      | wags |               |                  | true              | true               |                   |
 
     And user 3 has the following notification rules:
       | entities | tags            | warning_media | critical_media   | warning_blackhole | critical_blackhole | time_restrictions |
@@ -52,6 +53,7 @@ Feature: Notification rules on a per contact basis
       | baz      |                 | sms           | sms              |                   |                    |                   |
       | buf      |                 | email         | email            |                   |                    |                   |
       | buf      |                 | sms           | sms              |                   |                    |                   |
+      | bar      |                 | email         | email            | true              | true               |                   |
 
     And user 4 has the following notification rules:
       | entities | tags            | warning_media | critical_media   | warning_blackhole | critical_blackhole | time_restrictions |
@@ -151,8 +153,38 @@ Feature: Notification rules on a per contact basis
     And   an ok event is received
     Then  2 email alert should be queued for malak@example.com
 
-  @blackhole
-  Scenario: Drop alerts matching a blackhole rule
+  @blackhole @time
+  Scenario: Drop alerts matching a general blackhole rule
+    Given the check is check 'ping' on entity 'buf'
+    And   the check is in an ok state
+    When  a critical event is received
+    And   1 minute passes
+    And   a critical event is received
+    Then  0 email alerts should be queued for malak@example.com
+
+  @blackhole @time
+  Scenario: Drop alerts matching a blackhole rule by entity
+    Given the check is check 'ping' on entity 'bar'
+    And   the check is in an ok state
+    When  a warning event is received
+    And   1 minute passes
+    And   a warning event is received
+    Then  0 email alerts should be queued for malak@example.com
+    And   0 email alerts should be queued for vera@example.com
+    When  an ok event is received
+    Then  0 email alerts should be queued for malak@example.com
+    And   0 email alerts should be queued for vera@example.com
+
+  @blackhole @time
+  Scenario: Drop alerts matching a blackhole rule by tags
+    Given the check is check 'wags the dog' on entity 'bar'
+    And   the check is in an ok state
+    When  a warning event is received
+    And   1 minute passes
+    And   a warning event is received
+    Then  0 email alerts should be queued for imani@example.com
+    When  an ok event is received
+    Then  0 email alerts should be queued for imani@example.com
 
   @intervals @time
   Scenario: Alerts according to custom interval
@@ -274,7 +306,7 @@ Feature: Notification rules on a per contact basis
 
   @time
   Scenario: Test notifications behave like a critical notification
-    Given the check is check 'ping' on entity 'foo'
+    Given the check is check 'ping' on entity 'baz'
     And   the check is in an ok state
     When  a test event is received
     Then  1 email alert should be queued for malak@example.com
@@ -296,7 +328,7 @@ Feature: Notification rules on a per contact basis
     And   2 sms alert should be queued for +61400000001
 
   Scenario: Unknown event during unscheduled maintenance
-    Given the check is check 'ping' on entity 'foo'
+    Given the check is check 'ping' on entity 'baz'
     And   the check is in an ok state
     When  an unknown event is received
     And   1 minute passes

data/features/steps/notifications_steps.rb

@@ -139,6 +139,7 @@ When /^the SMS notification handler runs successfully$/ do
   @request = stub_request(:get, /^#{Regexp.escape(Flapjack::Gateways::SmsMessagenet::MESSAGENET_URL)}/)
 
   Flapjack::Gateways::SmsMessagenet.instance_variable_set('@config', {'username' => 'abcd', 'password' => 'efgh'})
+  Flapjack::Gateways::SmsMessagenet.instance_variable_set('@redis', @redis)
   Flapjack::Gateways::SmsMessagenet.instance_variable_set('@logger', @logger)
   Flapjack::Gateways::SmsMessagenet.start
 
@@ -148,6 +149,7 @@ end
 When /^the SMS notification handler fails to send an SMS$/ do
   @request = stub_request(:get, /^#{Regexp.escape(Flapjack::Gateways::SmsMessagenet::MESSAGENET_URL)}/).to_return(:status => [500, "Internal Server Error"])
   Flapjack::Gateways::SmsMessagenet.instance_variable_set('@config', {'username' => 'abcd', 'password' => 'efgh'})
+  Flapjack::Gateways::SmsMessagenet.instance_variable_set('@redis', @redis)
   Flapjack::Gateways::SmsMessagenet.instance_variable_set('@logger', @logger)
   Flapjack::Gateways::SmsMessagenet.start
 
@@ -157,6 +159,7 @@ end
 When /^the email notification handler runs successfully$/ do
   Resque.redis = @redis
   Flapjack::Gateways::Email.instance_variable_set('@config', {'smtp_config' => {'host' => '127.0.0.1', 'port' => 2525}})
+  Flapjack::Gateways::Email.instance_variable_set('@redis', @redis)
   Flapjack::Gateways::Email.instance_variable_set('@logger', @logger)
   Flapjack::Gateways::Email.start
 
@@ -175,6 +178,7 @@ end
 When /^the email notification handler fails to send an email$/ do
   Resque.redis = @redis
   Flapjack::Gateways::Email.instance_variable_set('@config', {'smtp_config' => {'host' => '127.0.0.1', 'port' => 2525}})
+  Flapjack::Gateways::Email.instance_variable_set('@redis', @redis)
   Flapjack::Gateways::Email.instance_variable_set('@logger', @logger)
   Flapjack::Gateways::Email.start
 

data/flapjack.gemspec

@@ -33,9 +33,10 @@ Gem::Specification.new do |gem|
   gem.add_dependency 'blather', '~> 0.8.3'
   gem.add_dependency 'chronic'
   gem.add_dependency 'chronic_duration'
-  gem.add_dependency 'activesupport'
+  gem.add_dependency 'activesupport', '~> 3.2.14'
   gem.add_dependency 'ice_cube'
-  gem.add_dependency 'tzinfo'
+  gem.add_dependency 'tzinfo', '~> 1.0.1'
+  gem.add_dependency 'tzinfo-data'
 
   gem.add_development_dependency 'rake'
 end

data/lib/flapjack/coordinator.rb

@@ -31,11 +31,13 @@ module Flapjack
         setup_signals if options[:signals]
         add_pikelets(pikelets(@config.all))
       end
+
+      @exit_value
     end
 
-    def stop
-      return if @stopping
-      @stopping = true
+    def stop(value = 0)
+      return unless @exit_value.nil?
+      @exit_value = value
       remove_pikelets(@pikelets, :shutdown => true)
       # Syslog.close if Syslog.opened? # TODO revisit in threading branch
     end
@@ -111,10 +113,10 @@ module Flapjack
     # within a single coordinator instance. Coordinator is essentially
     # a singleton anyway...
     def setup_signals
-      Kernel.trap('INT')  { stop }
-      Kernel.trap('TERM') { stop }
+      Kernel.trap('INT')  { stop(Signal.list['INT']) }
+      Kernel.trap('TERM') { stop(Signal.list['TERM']) }
       unless RbConfig::CONFIG['host_os'] =~ /mswin|windows|cygwin/i
-        Kernel.trap('QUIT') { stop }
+        Kernel.trap('QUIT') { stop(Signal.list['QUIT']) }
         Kernel.trap('HUP')  { reload }
       end
     end