cloudtasker 0.11.0 → 0.12.rc1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8f4e53fe9c99a8c6a74e064d7b7c562cdb671682aaed9bb6112a77c7fd453124
-  data.tar.gz: 3b2e6aeca4fde352acd7b0db6fe469221514bf7fd753c07c3d9540d881bf1c2f
+  metadata.gz: b89246a4e0d7f658fdb3115c3a00b54df0e55b930428fe52caef471386119906
+  data.tar.gz: 4fe22aa2588e95fffcb223e32e50a2baa5df2730d1e04f8f9d0202aa4138fbd4
 SHA512:
-  metadata.gz: 4be2ce9466e7ffd2cc274d57d79eedad96ae56d699632f88bfdf6384413aa9bfa19dff8a124926bd2b41229ce53f3126b2bd04646cc9125136bc93fdffb0642a
-  data.tar.gz: f6a3f7fa152f5ebedbe25945e86538691ee658c5bc02b840fd3d1f337ba0fcd78db6a84c52bcd6ade01af51335ee81d457152bfa9767d3ac9541b5d31ef48126
+  metadata.gz: 1ea18e7601bb394ad4d7f72838bf3eaa9d7ff05df0f0b4f3d08e5d7fa82cee52e51be0c4148d0ac282986f366f6159d0bcc634993f07fbbf28c63ec021d29b5e
+  data.tar.gz: 9282074934628583c6c7b5d2bc6efaf6fac87777bab718adf2204f3eb87143f8bce950711395f6d1ecc6dac6fb349fef52bc94eeb42352322d824b6b35011d5a

data/.rubocop.yml

@@ -13,6 +13,8 @@ Metrics/ModuleLength:
 
 Metrics/AbcSize:
   Max: 20
+  Exclude:
+    - 'spec/support/*'
 
 Metrics/LineLength:
   Max: 120
@@ -47,4 +49,9 @@ Metrics/ParameterLists:
   CountKeywordArgs: false
 
 RSpec/MessageSpies:
-  Enabled: false
+  Enabled: false
+
+RSpec/MultipleExpectations:
+  Exclude:
+    - 'examples/**/*'
+    - 'spec/integration/**/*'

data/CHANGELOG.md

@@ -1,6 +1,20 @@
 # Changelog
 
-## [v0.11.0](https://github.com/keypup-io/cloudtasker/tree/v0.11.0) (2020-11-23)
+## Latest RC [v0.12.rc1](https://github.com/keypup-io/cloudtasker/tree/v0.12.rc1) (2021-03-11)
+
+[Full Changelog](https://github.com/keypup-io/cloudtasker/compare/v0.11.0...v0.12.rc1)
+
+**Improvements:**
+- ActiveJob: do not double log errors (ActiveJob has its own error logging)
+- Error logging: Use worker logger so as to include context (job args etc.)
+- Error logging: Do not log exception and stack trace separately, combine them instead.
+- Batch callbacks: Retry jobs when completion callback fails
+- Redis: Use Redis Sets instead of key pattern matching for listing methods (Cron jobs and Local Server)
+
+**Fixed bugs:**
+- Retries: Enforce job retry limit on job processing. There was an edge case where jobs could be retried indefinitely on batch callback errors.
+
+## [v0.11.0](https://github.com/keypup-io/cloudtasker/tree/v0.11.0) (2021-03-11)
 
 [Full Changelog](https://github.com/keypup-io/cloudtasker/compare/v0.10.0...v0.11.0)
 

data/app/controllers/cloudtasker/worker_controller.rb

@@ -19,32 +19,19 @@ module Cloudtasker
       # Process payload
       WorkerHandler.execute_from_payload!(payload)
       head :no_content
-    rescue DeadWorkerError, MissingWorkerArgumentsError => e
+    rescue DeadWorkerError, MissingWorkerArgumentsError
       # 205: job will NOT be retried
-      log_error(e)
       head :reset_content
-    rescue InvalidWorkerError => e
+    rescue InvalidWorkerError
       # 404: Job will be retried
-      log_error(e)
       head :not_found
-    rescue StandardError => e
-      # 404: Job will be retried
-      log_error(e)
+    rescue StandardError
+      # 422: Job will be retried
       head :unprocessable_entity
     end
 
     private
 
-    #
-    # Log an error via cloudtasker logger.
-    #
-    # @param [Exception] e The error to log
-    #
-    def log_error(error)
-      Cloudtasker.logger.error(error)
-      Cloudtasker.logger.error(error.backtrace.join("\n"))
-    end
-
     #
     # Parse the request body and return the actual job
     # payload.

data/lib/cloudtasker/backend/redis_task.rb

@@ -23,14 +23,12 @@ module Cloudtasker
       #
       # Return a namespaced key.
       #
-      # @param [String, Symbol] val The key to namespace
+      # @param [String, Symbol, nil] val The key to namespace
       #
       # @return [String] The namespaced key.
       #
-      def self.key(val)
-        return nil if val.nil?
-
-        [to_s.underscore, val.to_s].join('/')
+      def self.key(val = nil)
+        [to_s.underscore, val].compact.map(&:to_s).join('/')
       end
 
       #
@@ -39,9 +37,17 @@ module Cloudtasker
       # @return [Array<Cloudtasker::Backend::RedisTask>] All the tasks.
       #
       def self.all
-        redis.search(key('*')).map do |gid|
-          payload = redis.fetch(gid)
-          new(payload.merge(id: gid.sub(key(''), '')))
+        if redis.exists?(key)
+          # Use Schedule Set if available
+          redis.smembers(key).map { |id| find(id) }
+        else
+          # Fallback to redis key matching and migrate tasks
+          # to use Task Set instead.
+          redis.search(key('*')).map do |gid|
+            task_id = gid.sub(key(''), '')
+            redis.sadd(key, task_id)
+            find(task_id)
+          end
         end
       end
 
@@ -82,6 +88,7 @@ module Cloudtasker
 
         # Save job
         redis.write(key(id), payload)
+        redis.sadd(key, id)
         new(payload.merge(id: id))
       end
 
@@ -105,6 +112,7 @@ module Cloudtasker
       # @param [String] id The task id.
       #
       def self.delete(id)
+        redis.srem(key, id)
         redis.del(key(id))
       end
 
@@ -176,7 +184,7 @@ module Cloudtasker
       # Remove the task from the queue.
       #
       def destroy
-        redis.del(gid)
+        self.class.delete(id)
       end
 
       #

data/lib/cloudtasker/batch/job.rb

@@ -13,6 +13,10 @@ module Cloudtasker
       # List of statuses triggering a completion callback
       COMPLETION_STATUSES = %w[completed dead].freeze
 
+      # These callbacks do not need to raise errors on their own
+      # because the jobs will be either retried or dropped
+      IGNORED_ERRORED_CALLBACKS = %i[on_child_error on_child_dead].freeze
+
       #
       # Return the cloudtasker redis client
       #
@@ -250,8 +254,8 @@ module Cloudtasker
       end
 
       #
-      # Run worker callback in a controlled environment to
-      # avoid interruption of the callback flow.
+      # Run worker callback. The error and dead callbacks get
+      # silenced should they raise an error.
       #
       # @param [String, Symbol] callback The callback to run.
       # @param [Array<any>] *args The callback arguments.
@@ -261,9 +265,15 @@ module Cloudtasker
       def run_worker_callback(callback, *args)
         worker.try(callback, *args)
       rescue StandardError => e
-        Cloudtasker.logger.error("Error running callback #{callback}: #{e}")
-        Cloudtasker.logger.error(e.backtrace.join("\n"))
-        nil
+        # There is no point in retrying jobs due to failure callbacks failing
+        # Only completion callbacks will trigger a re-run of the job because
+        # these do matter for batch completion
+        raise(e) unless IGNORED_ERRORED_CALLBACKS.include?(callback)
+
+        # Log error instead
+        worker.logger.error(
+          ["Callback #{callback} failed to run. Skipping to preserve error flow.", e, e.backtrace].flatten.join("\n")
+        )
       end
 
       #
@@ -275,7 +285,7 @@ module Cloudtasker
 
         # Propagate event
         parent_batch&.on_child_complete(self, status)
-      ensure
+
         # The batch tree is complete. Cleanup the tree.
         cleanup unless parent_batch
       end

data/lib/cloudtasker/cron/schedule.rb

@@ -21,14 +21,12 @@ module Cloudtasker
       #
       # Return a namespaced key.
       #
-      # @param [String, Symbol] val The key to namespace
+      # @param [String, Symbol, nil] val The key to namespace
       #
       # @return [String] The namespaced key.
       #
-      def self.key(val)
-        return nil if val.nil?
-
-        [to_s.underscore, val.to_s].join('/')
+      def self.key(val = nil)
+        [to_s.underscore, val].compact.map(&:to_s).join('/')
       end
 
       #
@@ -37,8 +35,17 @@ module Cloudtasker
       # @return [Array<Cloudtasker::Batch::Schedule>] The list of stored schedules.
       #
       def self.all
-        redis.search(key('*')).map do |gid|
-          find(gid.sub(key(''), ''))
+        if redis.exists?(key)
+          # Use Schedule Set if available
+          redis.smembers(key).map { |id| find(id) }
+        else
+          # Fallback to redis key matching and migrate schedules
+          # to use Schedule Set instead.
+          redis.search(key('*')).map do |gid|
+            schedule_id = gid.sub(key(''), '')
+            redis.sadd(key, schedule_id)
+            find(schedule_id)
+          end
         end
       end
 
@@ -90,7 +97,7 @@ module Cloudtasker
       end
 
       #
-      # Destroy a schedule by id.
+      # Delete a schedule by id.
       #
       # @param [String] id The schedule id.
       #
@@ -101,6 +108,7 @@ module Cloudtasker
 
           # Delete task and stored schedule
           CloudTask.delete(schedule.task_id) if schedule.task_id
+          redis.srem(key, schedule.id)
           redis.del(schedule.gid)
         end
       end
@@ -270,6 +278,7 @@ module Cloudtasker
 
         # Save schedule
         config_was_changed = config_changed?
+        redis.sadd(self.class.key, id)
         redis.write(gid, to_h)
 
         # Stop there if backend does not need update

data/lib/cloudtasker/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Cloudtasker
-  VERSION = '0.11.0'
+  VERSION = '0.12.rc1'
 end

data/lib/cloudtasker/worker.rb

@@ -311,13 +311,25 @@ module Cloudtasker
     end
 
     #
-    # Return true if the job has excceeded its maximum number
-    # of retries
+    # Return true if the job must declared dead upon raising
+    # an error.
+    #
+    # @return [Boolean] True if the job must die on error.
+    #
+    def job_must_die?
+      job_retries >= job_max_retries
+    end
+
+    #
+    # Return true if the job has strictly excceeded its maximum number
+    # of retries.
+    #
+    # Used a preemptive filter when running the job.
     #
     # @return [Boolean] True if the job is dead
     #
     def job_dead?
-      job_retries >= job_max_retries
+      job_retries > job_max_retries
     end
 
     #
@@ -332,11 +344,35 @@ module Cloudtasker
       (perform_ended_at - perform_started_at).ceil(3)
     end
 
+    #
+    # Run worker callback.
+    #
+    # @param [String, Symbol] callback The callback to run.
+    # @param [Array<any>] *args The callback arguments.
+    #
+    # @return [any] The callback return value
+    #
+    def run_callback(callback, *args)
+      try(callback, *args)
+    end
+
     #=============================
     # Private
     #=============================
     private
 
+    #
+    # Flag the worker as dead by invoking the on_dead hook
+    # and raising a DeadWorkerError
+    #
+    # @param [Exception, nil] error An optional exception to be passed to the DeadWorkerError.
+    #
+    def flag_as_dead(error = nil)
+      run_callback(:on_dead, error || DeadWorkerError.new)
+    ensure
+      raise(DeadWorkerError, error)
+    end
+
     #
     # Execute the worker perform method through the middleware chain.
     #
@@ -346,6 +382,9 @@ module Cloudtasker
       self.perform_started_at = Time.now
 
       Cloudtasker.config.server_middleware.invoke(self) do
+        # Immediately abort the job if it is already dead
+        flag_as_dead if job_dead?
+
         begin
           # Abort if arguments are missing. This may happen with redis arguments storage
           # if Cloud Tasks times out on a job but the job still succeeds
@@ -356,12 +395,11 @@ module Cloudtasker
           # Perform the job
           perform(*job_args)
         rescue StandardError => e
-          try(:on_error, e)
-          return raise(e) unless job_dead?
+          run_callback(:on_error, e)
+          return raise(e) unless job_must_die?
 
           # Flag job as dead
-          try(:on_dead, e)
-          raise(DeadWorkerError, e)
+          flag_as_dead(e)
         end
       end
     ensure

data/lib/cloudtasker/worker_handler.rb

@@ -45,6 +45,28 @@ module Cloudtasker
       end
     end
 
+    #
+    # Log error on execution failure.
+    #
+    # @param [Cloudtasker::Worker, nil] worker The worker.
+    # @param [Exception] error The error to log.
+    #
+    # @void
+    #
+    def self.log_execution_error(worker, error)
+      # ActiveJob has its own error logging. No need to double log the error.
+      # Note: we use string matching instead of class matching as
+      # ActiveJob::QueueAdapters::CloudtaskerAdapter::JobWrapper might not be loaded
+      return if worker.class.to_s =~ /^ActiveJob::/
+
+      # Choose logger to use based on context
+      # Worker will be nil on InvalidWorkerError - in that case we use generic logging
+      logger = worker&.logger || Cloudtasker.logger
+
+      # Log error
+      logger.error([error, error.backtrace].flatten.join("\n"))
+    end
+
     #
     # Execute a task worker from a task payload
     #
@@ -88,6 +110,10 @@ module Cloudtasker
     rescue DeadWorkerError, MissingWorkerArgumentsError => e
       # Delete stored args payload if job is dead
       redis.expire(args_payload_key, ARGS_PAYLOAD_CLEANUP_TTL) if args_payload_key
+      log_execution_error(worker, e)
+      raise(e)
+    rescue StandardError => e
+      log_execution_error(worker, e)
       raise(e)
     end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: cloudtasker
 version: !ruby/object:Gem::Version
-  version: 0.11.0
+  version: 0.12.rc1
 platform: ruby
 authors:
 - Arnaud Lachaume
@@ -427,9 +427,9 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - ">"
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 1.3.1
 requirements: []
 rubygems_version: 3.0.0
 signing_key: