flapjack 1.2.0 → 1.2.1rc1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 04d31109d0b39ced29a1b027a0894cab07b9d4c0
-  data.tar.gz: 86e20fc3b3f0d34acca6be7e33468fc3d609e868
+  metadata.gz: 1038ee80ed4a85514dbe214c5a50768c3517db8e
+  data.tar.gz: 6373cd02ed2d57411ed732bb1b00eb3e449b6683
 SHA512:
-  metadata.gz: a838f02684a7cee3ba40cc5be1ecfa2e3b15246912a7bff218944b323f14d43ef64ebce1e9ce978f37795b17d44acc4ec708f4a8768e9788df339a20619f1892
-  data.tar.gz: 93784a164826ef9af3bf84b947d0debdbf21acf3e601ee809c248534d52e0af2cbea643e5cd7ab2c4af4e14a7b8b6ced725883739ecd148ae11f3492a4771f4a
+  metadata.gz: 4ebdf21010802ae6ff6cd737e0ca6be03428fbafd6cc7dd31a65da6fc8fe821c3bd35a375e5a2b0a05ea92a438fdd0e2909b5f6f65ff69a99d1eb5daabc78c7c
+  data.tar.gz: 91d761e805b496c781596559c1d32465c875208db46365c31f60e04b19c89ef34faf3d3e0adac3425149d02421433699b8f1b1adcd3fd0b19437f063bbcbe3ba

data/CHANGELOG.md

@@ -1,5 +1,15 @@
 ## Flapjack Changelog
 
+# 1.2.1rc1 - 2014-12-04
+- Bug: Exception in pagerduty gateway when looking for acknowledgements #721 (@jessereynolds)
+- Bug: Retrieving scheduled maintenance via api fails when old entities without ids are present #711 (@jessereynolds)
+- Bug: Repeated acknowledgement notifications sent after acknowledging on PagerDuty #714 (@jessereynolds)
+- Bug: Exception thrown to browser when you pass no values to "Add Scheduled Maintenance" #719 (@ali-graham)
+- Bug: API doesn't allow contact timezone to be updated #718 (@ali-graham)
+- Bug: IceCube deprecation warning in 1.2.0 - :exrules is deprecated, #715 (@ali-graham)
+- Bug: Badly formed error response to scheduled maintenance report with invalid check id #712 (@ali-graham)
+- Bug: Searching for checks in maintenance via CLI doesn't seem to work #710 (@jessereynolds, @Hobbsee)
+
 # 1.2.0 - 2014-11-07
 - Bug: multi blocks for safe redis connection pool usage #694 (@ali-graham)
 - Bug: data migration to work around previous notification rule bug #699 (@ali-graham)

data/Gemfile-ruby1.9.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    flapjack (1.2.0)
+    flapjack (1.2.1rc1)
       activesupport (~> 3.2.14)
       blather (~> 0.8.3)
       chronic

data/etc/flapjack_config.yaml.example

@@ -172,6 +172,7 @@ production:
       api_url: "http://localhost:3081/"
       # Full path to location of logo file, e.g. /etc/flapjack/custom_logo.png
       #logo_image_path: "/etc/flapjack/web/custom_logo/flapjack-2013-notext-transparent-300-300.png"
+      show_exceptions: false
       logger:
         level: INFO
         syslog_errors: yes
@@ -376,6 +377,7 @@ development:
       api_url: "http://localhost:3081/"
       # Full path to location of logo file, e.g. /etc/flapjack/custom_logo.png
       #logo_image_path: "/etc/flapjack/web/custom_logo/flapjack-2013-notext-transparent-300-300.png"
+      show_exceptions: true
       logger:
         level: DEBUG
         syslog_errors: yes

data/features/ack_after_sched_maint.feature

@@ -0,0 +1,44 @@
+@rollup @notification_rules @resque @processor @notifier @events
+Feature: Multiple acknowledgements after scheduled maintenance
+
+  Background:
+    Given the following users exist:
+      | id  | first_name | last_name | email             | sms          | timezone         |
+      | 1   | Malak      | Al-Musawi | malak@example.com | +61400000001 | Asia/Baghdad     |
+
+    And the following entities exist:
+      | id  | name           | contacts |
+      | 1   | foo            | 1        |
+
+    And user 1 has the following notification intervals:
+      | email | sms |
+      | 15    | 15  |
+
+    And user 1 has the following notification rollup thresholds:
+      | email | sms |
+      | 3     | 3   |
+
+    And user 1 has the following notification rules:
+      | entities | unknown_media | warning_media | critical_media   |
+      |          |               | email         | sms,email        |
+
+  @time
+  Scenario: Multiple acks after sched maint ends
+    Given the check is check 'ping' on entity 'foo'
+    And   the check is in an ok state
+    And   the check is in scheduled maintenance for 1 hour
+    When  1 minute passes
+    And   a critical event is received
+    Then  no email alerts should be queued for malak@example.com
+    When  60 minutes passes
+    And   a critical event is received
+    Then  1 email alert of type problem should be queued for malak@example.com
+    And   the check should appear in unacknowledged_failing
+    When  1 minute passes
+    And   an acknowledgement event is received
+    Then  1 email alert of type acknowledgement should be queued for malak@example.com
+    And   the check should not appear in unacknowledged_failing
+    When  1 minute passes
+    And   an acknowledgement event is received
+    Then  2 email alert of type acknowledgement should be queued for malak@example.com
+

data/features/events.feature

@@ -199,6 +199,13 @@ Feature: events
     When  an acknowledgement event is received
     Then  a notification should be generated
 
+  Scenario: Acknowledgement when acknowledged
+    Given the check is in a critical state
+    When  an acknowledgement event is received
+    Then  a notification should be generated
+    When  an acknowledgement event is received
+    Then  a notification should be generated
+
   Scenario: Brief critical then OK
     Given the check is in an ok state
     When  a critical event is received

data/features/steps/events_steps.rb

@@ -426,3 +426,17 @@ When(/^user (\S+) ceases to be a contact of entity '(.*)'$/) do |contact_id, ent
   @redis.srem("contacts_for:#{entity.id}", contact_id)
 end
 
+Then(/^(?:the check|check '([\w\.\-]+)' for entity '([\w\.\-]+)') should not appear in unacknowledged_failing$/) do |check, entity|
+  check  ||= @check
+  entity ||= @entity
+  unacknowledged_failing_checks = Flapjack::Data::EntityCheck.unacknowledged_failing(:redis => @redis)
+  expect(unacknowledged_failing_checks.map {|ec| "#{ec.entity.name}:#{ec.check}"}).to_not include("#{entity}:#{check}")
+end
+
+Then(/^(?:the check|check '([\w\.\-]+)' for entity '([\w\.\-]+)') should appear in unacknowledged_failing$/) do |check, entity|
+  check  ||= @check
+  entity ||= @entity
+  unacknowledged_failing_checks = Flapjack::Data::EntityCheck.unacknowledged_failing(:redis => @redis)
+  expect(unacknowledged_failing_checks.map {|ec| "#{ec.entity.name}:#{ec.check}"}).to include("#{entity}:#{check}")
+end
+

data/lib/flapjack/cli/maintenance.rb

@@ -57,7 +57,6 @@ module Flapjack
       end
 
       def create
-        exit_now!("Entity & check must be supplied to create a maintenance period") if @options[:entity].nil? || @options[:check].nil?
         errors = Flapjack::Data::EntityCheck.create_maintenance(@options)
         (errors.each { |k, v| puts "#{k}: #{v}" }; exit_now!('Failed to create maintenances')) if errors.length > 0
         puts "The maintenances specified have been created"
@@ -83,28 +82,28 @@ command :maintenance do |maintenance|
   maintenance.desc 'Show maintenance windows according to criteria (default: all ongoing maintenance)'
   maintenance.command :show do |show|
 
-    show.flag [:e, 'entity'],
+    show.flag ['entity', :e],
       :desc => 'The entity for the maintenance window to occur on.  This can be a string, or a ruby regex of the form \'db*\' or \'[[:lower:]]\''
 
-    show.flag [:c, 'check'],
+    show.flag ['check', :c],
       :desc => 'The check for the maintenance window to occur on.  This can be a string, or a ruby regex of the form \'http*\' or \'[[:lower:]]\''
 
-    show.flag [:r, 'reason'],
+    show.flag ['reason', :r],
       :desc => 'The reason for the maintenance window to occur.  This can be a string, or a ruby regex of the form \'Downtime for *\' or \'[[:lower:]]\''
 
-    show.flag [:s, 'start', 'started', 'starting'],
-      :desc => 'The start time for the maintenance window. This should be prefixed with "more than", "less than", "on", "before", or "after", or of the form "between times and time"'
+    show.flag ['start', 'started', 'starting', :s],
+      :desc => 'The start time for the maintenance window, eg \'before 10am\'. This should be prefixed with \'more than\', \'less than\', \'on\', \'before\', or \'after\', or of the form \'between times and time\''
 
-    show.flag [:d, 'duration'],
-      :desc => 'The total duration of the maintenance window. This should be prefixed with "more than", "less than", "before, "after" or "equal to", or or of the form "between 3 and 4 hours".  This should be an interval'
+    show.flag ['duration', :d],
+      :desc => 'The total duration of the maintenance window, eg \'equal to 5 hours\'. This should be prefixed with \'more than\', \'less than\', \'before, \'after\' or \'equal to\', or or of the form \'between 3 and 4 hours\'.  This should be an interval'
 
-    show.flag [:f, 'finish', 'finished', 'finishing', 'remain', 'remained', 'remaining', 'end'],
-      :desc => 'The finishing time for the maintenance window. This should be prefixed with "more than", "less than", "on", "before", or "after", or of the form "between time and time"'
+    show.flag ['finish', 'finished', 'finishing', 'remain', 'remained', 'remaining', 'end', :f],
+      :desc => 'The finishing time for the maintenance window, eg \'more than 1 year from now\'. This should be prefixed with \'more than\', \'less than\', \'on\', \'before\', or \'after\', or of the form \'between time and time\''
 
-    show.flag [:st, 'state'],
+    show.flag ['state', :st],
       :desc => 'The state that the check is currently in'
 
-    show.flag [:t, 'type'],
+    show.flag ['type', :t],
       :desc => 'The type of maintenance scheduled',
       :default_value => 'scheduled'
 
@@ -117,32 +116,32 @@ command :maintenance do |maintenance|
   maintenance.desc 'Delete maintenance windows according to criteria (default: all ongoing maintenance)'
   maintenance.command :delete do |delete|
 
-    delete.flag [:a, 'apply'],
+    delete.flag ['apply', :a],
       :desc => 'Whether this deletion should occur',
       :default_value => false
 
-    delete.flag [:e, 'entity'],
+    delete.flag ['entity', :e],
       :desc => 'The entity for the maintenance window to occur on.  This can be a string, or a ruby regex of the form \'db*\' or \'[[:lower:]]\''
 
-    delete.flag [:c, 'check'],
+    delete.flag ['check', :c],
       :desc => 'The check for the maintenance window to occur on.  This can be a string, or a ruby regex of the form \'http*\' or \'[[:lower:]]\''
 
-    delete.flag [:r, 'reason'],
+    delete.flag ['reason', :r],
       :desc => 'The reason for the maintenance window to occur.  This can be a string, or a ruby regex of the form \'Downtime for *\' or \'[[:lower:]]\''
 
-    delete.flag [:s, 'start', 'started', 'starting'],
-      :desc => 'The start time for the maintenance window. This should be prefixed with "more than", "less than", "on", "before", or "after", or of the form "between times and time"'
+    delete.flag ['start', 'started', 'starting', :s],
+      :desc => 'The start time for the maintenance window, eg \'before 10am\'. This should be prefixed with \'more than\', \'less than\', \'on\', \'before\', or \'after\', or of the form \'between times and time\''
 
-    delete.flag [:d, 'duration'],
-      :desc => 'The total duration of the maintenance window. This should be prefixed with "more than", "less than", "before, "after" or "equal to", or or of the form "between 3 and 4 hours".  This should be an interval'
+    delete.flag ['duration', :d],
+      :desc => 'The total duration of the maintenance window, eg \'equal to 5 hours\'. This should be prefixed with \'more than\', \'less than\', \'before, \'after\' or \'equal to\', or or of the form \'between 3 and 4 hours\'.  This should be an interval'
 
-    delete.flag [:f, 'finish', 'finished', 'finishing', 'remain', 'remained', 'remaining', 'end'],
-      :desc => 'The finishing time for the maintenance window. This should be prefixed with "more than", "less than", "on", "before", or "after", or of the form "between time and time"'
+    delete.flag ['finish', 'finished', 'finishing', 'remain', 'remained', 'remaining', 'end', :f],
+      :desc => 'The finishing time for the maintenance window, eg \'more than 1 year from now\'. This should be prefixed with \'more than\', \'less than\', \'on\', \'before\', or \'after\', or of the form \'between time and time\''
 
-    delete.flag [:st, 'state'],
+    delete.flag ['state', :st],
       :desc => 'The state that the check is currently in'
 
-    delete.flag [:t, 'type'],
+    delete.flag ['type', :t],
       :desc => 'The type of maintenance scheduled',
       :default_value => 'scheduled'
 
@@ -155,25 +154,30 @@ command :maintenance do |maintenance|
   maintenance.desc 'Create a maintenance window'
   maintenance.command :create do |create|
 
-    create.flag [:e, 'entity'],
+    create.flag ['entity', :e],
       :desc => 'The entity for the maintenance window to occur on.  This can be a comma separated list',
-      :type => Array
+      :type => Array,
+      :required => true
 
-    create.flag [:c, 'check'],
+    create.flag ['check', :c],
       :desc => 'The check for the maintenance window to occur on.  This can be a comma separated list',
-      :type => Array
+      :type => Array,
+      :required => true
 
-    create.flag [:r, 'reason'],
-      :desc => 'The reason for the maintenance window to occur'
+    create.flag ['reason', :r],
+      :desc => 'The reason for the maintenance window to occur',
+      :required => true
 
-    create.flag [:s, 'start', 'started', 'starting'],
-      :desc => 'The start time for the maintenance window'
+    create.flag ['start', 'started', 'starting', :s],
+      :desc => 'The start time for the maintenance window, eg \'now\' or \'in 3 hours\'',
+      :required => true
 
-    create.flag [:d, 'duration'],
-      :desc => 'The total duration of the maintenance window.  This should be an interval'
+    create.flag ['duration', :d],
+      :desc => 'The total duration of the maintenance window, eg \'30 minutes\'.  This should be an interval',
+      :required => true
 
-    create.flag [:t, 'type'],
-      :desc => 'The type of maintenance scheduled ("scheduled")',
+    create.flag ['type', :t],
+      :desc => 'The type of maintenance scheduled (\'scheduled\')',
       :default_value => 'scheduled'
 
     create.action do |global_options,options,args|

data/lib/flapjack/data/entity_check.rb

@@ -160,7 +160,7 @@ module Flapjack
           redis.exists(entity_check + ':unscheduled_maintenance')
         }.collect {|entity_check|
           Flapjack::Data::EntityCheck.for_event_id(entity_check, :redis => redis)
-        }
+        }.compact
       end
 
       def self.find_maintenance(options = {})
@@ -983,8 +983,8 @@ module Flapjack
 
       def initialize(entity, check, options = {})
         raise "Redis connection not set" unless @redis = options[:redis]
-        raise "Invalid entity" unless @entity = entity
-        raise "Invalid check" unless @check = check
+        raise "Invalid entity (#{entity.inspect})" unless @entity = entity
+        raise "Invalid check (#{check.inspect} on #{entity.inspect})" unless @check = check
         @key = "#{entity.name}:#{check}"
         if @redis.zscore("all_checks", @key).nil?
           timestamp = options[:timestamp] || Time.now.to_i

data/lib/flapjack/data/migration.rb

@@ -18,10 +18,10 @@ module Flapjack
         semaphore = nil
         strikes = 0
         begin
-          semaphore = Flapjack::Data::Semaphore.new(resource, :redis => redis, :expiry => 60)
+          semaphore = Flapjack::Data::Semaphore.new(resource, :redis => redis, :expiry => 300)
         rescue Flapjack::Data::Semaphore::ResourceLocked
           strikes += 1
-          if strikes < 5
+          if strikes < 10
             sleep 2
             retry
           end
@@ -30,66 +30,102 @@ module Flapjack
         semaphore
       end
 
-      def self.migrate_entity_check_data_if_required(options = {})
+      def self.create_entity_ids_if_required(options = {})
         raise "Redis connection not set" unless redis = options[:redis]
-
         logger = options[:logger]
 
+        if redis.exists('created_ids_for_old_entities_without_ids')
+          return
+        end
+
         semaphore = obtain_semaphore(ENTITY_DATA_MIGRATION, :redis => redis)
         if semaphore.nil?
           unless logger.nil?
-            logger.fatal "Could not obtain lock for data migration. Ensure that " +
+            logger.fatal "Could not obtain lock for data migration (entity id creation). Ensure that " +
               "no other flapjack processes are running that might be executing " +
               "migrations, check logs for any exceptions, manually delete the " +
               "'#{ENTITY_DATA_MIGRATION}' key from your Flapjack Redis " +
               "database and try running Flapjack again."
           end
-          exit
+          raise "Unable to obtain semaphore #{ENTITY_DATA_MIGRATION}"
         end
 
-        if redis.exists('all_checks')
+        begin
+          logger.warn "Ensuring all entities have ids ..." unless logger.nil?
+
+          Flapjack::Data::EntityCheck.find_current_names_by_entity(:redis => redis, :logger => logger).keys.each {|entity_name|
+            entity = Flapjack::Data::Entity.find_by_name(entity_name, :create => true, :redis => redis, :logger => logger)
+          }
+
+          all_checks = Flapjack::Data::EntityCheck.all(:redis => redis, :logger => logger, :create_entity => true)
+
+          redis.set('created_ids_for_old_entities_without_ids', 'true')
+          logger.warn "Entity id creation complete."
+        ensure
           semaphore.release
-          return
         end
+      end
 
-        logger.warn "Upgrading Flapjack's entity/check Redis indexes..." unless logger.nil?
+      def self.migrate_entity_check_data_if_required(options = {})
+        raise "Redis connection not set" unless redis = options[:redis]
 
-        check_names = redis.keys('check:*').map {|c| c.sub(/^check:/, '') } |
-          Flapjack::Data::EntityCheck.find_current_names(:redis => redis)
+        logger = options[:logger]
 
-        unless check_names.empty?
-          timestamp = Time.now.to_i
+        if redis.exists('all_checks')
+          return
+        end
 
-          check_names.each do |ecn|
-            redis.zadd("all_checks", timestamp, ecn)
-            entity_name, check = ecn.split(':', 2)
-            redis.zadd("all_checks:#{entity_name}", timestamp, check)
-            # not deleting the check hashes, they store useful data
+        semaphore = obtain_semaphore(ENTITY_DATA_MIGRATION, :redis => redis)
+        if semaphore.nil?
+          unless logger.nil?
+            logger.fatal "Could not obtain lock for entity check data migration. Ensure that " +
+              "no other flapjack processes are running that might be executing " +
+              "migrations, check logs for any exceptions, manually delete the " +
+              "'#{ENTITY_DATA_MIGRATION}' key from your Flapjack Redis " +
+              "database and try running Flapjack again."
           end
+          raise "Unable to obtain semaphore #{ENTITY_DATA_MIGRATION}"
         end
 
-        logger.warn "Checks indexed." unless logger.nil?
-
-        entity_name_keys = redis.keys("entity_id:*")
-        unless entity_name_keys.empty?
-          ids = redis.mget(*entity_name_keys)
+        begin
+          logger.warn "Upgrading Flapjack's entity/check Redis indexes..." unless logger.nil?
 
-          entity_name_keys.each do |enk|
-            enk =~ /^entity_id:(.+)$/; entity_name = $1; entity_id = ids.shift
+          check_names = redis.keys('check:*').map {|c| c.sub(/^check:/, '') } |
+            Flapjack::Data::EntityCheck.find_current_names(:redis => redis)
 
-            redis.hset('all_entity_names_by_id', entity_id, entity_name)
-            redis.hset('all_entity_ids_by_name', entity_name, entity_id)
+          unless check_names.empty?
+            timestamp = Time.now.to_i
 
-            redis.del(enk)
-            redis.del("entity:#{entity_id}")
+            check_names.each do |ecn|
+              redis.zadd("all_checks", timestamp, ecn)
+              entity_name, check = ecn.split(':', 2)
+              redis.zadd("all_checks:#{entity_name}", timestamp, check)
+              # not deleting the check hashes, they store useful data
+            end
           end
-        end
 
-        logger.warn "Entities indexed." unless logger.nil?
+          logger.warn "Checks indexed." unless logger.nil?
+
+          entity_name_keys = redis.keys("entity_id:*")
+          unless entity_name_keys.empty?
+            ids = redis.mget(*entity_name_keys)
+
+            entity_name_keys.each do |enk|
+              enk =~ /^entity_id:(.+)$/; entity_name = $1; entity_id = ids.shift
 
-        semaphore.release
+              redis.hset('all_entity_names_by_id', entity_id, entity_name)
+              redis.hset('all_entity_ids_by_name', entity_name, entity_id)
 
-        logger.warn "Indexing complete." unless logger.nil?
+              redis.del(enk)
+              redis.del("entity:#{entity_id}")
+            end
+          end
+
+          logger.warn "Entities indexed." unless logger.nil?
+          logger.warn "Indexing complete." unless logger.nil?
+        ensure
+          semaphore.release
+        end
       end
 
       def self.refresh_archive_index(options = {})
@@ -147,4 +183,4 @@ module Flapjack
 
     end
   end
-end
+end

data/lib/flapjack/data/notification_rule.rb

@@ -253,6 +253,11 @@ module Flapjack
         return unless (tr.has_key?(:start_time) || tr.has_key?(:start_date)) &&
           (tr.has_key?(:end_time) || tr.has_key?(:end_date))
 
+        # exrules is deprecated in latest ice_cube, but may be stored in data
+        # serialised from earlier versions of the gem
+        # ( https://github.com/flapjack/flapjack/issues/715 )
+        tr.delete(:exrules)
+
         parsed_time = proc {|tr, field|
           if t = tr.delete(field)
             t = t.dup