canvas_sync 0.17.0 → 0.17.3.beta3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d8629b92b654035d7c4480345bbf938d385b186c66acd5b4e30f8d76d6dcb109
-  data.tar.gz: 0ef5bae856795745c2f8c8f845533148028076df62f76a395b315ae8b55679ff
+  metadata.gz: 54feabc8f9154b633a70a11d15bfeebaf54ab9d9eb24b9a636c27844728ca1a3
+  data.tar.gz: 0c8d2692a83ebb37a518f61d11fb2970b3204f3d68c990c721fa23a7365a91b6
 SHA512:
-  metadata.gz: c08b8308faa3af77ba3828ef122d20e08d2dcf9f5a57e26f2274f000c23fa7d870638e9c2c37256139db6e63d9bc21e7b1497d5b257696c11c224ec9c77ae92e
-  data.tar.gz: 63ddbbaf804dd4e547885c9817c7793646fae10c6f54d76bcf82643cc821d3f72c2a3b76a085e7720c1ec50360cd327f8498481c2d6d04cd829677d0d732259a
+  metadata.gz: 50efdad1e463d363463b72fa2fcc5581f877bdeb2c2440d688b10f451321996d95c7acd610d822fcc943d0220ed5520f4963de12341c1adb4238d131c5d8df04
+  data.tar.gz: 51d40bc898a58a0aab65a85d4c1820ca5da72bf22b4c11e0ee631ad2e815dc344dc4824a2ade24e7485ace7ad0d12692d71cc93054afcfc05269ac22a4f7b4c9

data/README.md

@@ -246,6 +246,64 @@ A batch can be created using `Sidekiq::Batch` or `CanvasSync::JobBatching::Batch
 
 Also see `canvas_sync/jobs/begin_sync_chain_job`, `canvas_sync/Job_batches/jobs/serial_batch_job`, or `canvas_sync/Job_batches/jobs/concurrent_batch_job` for example usage.
 
+Example:
+```ruby
+batch = CanvasSync::JobBatches::Batch.new
+batch.description = "Some Batch" # Optional, but can be useful for debugging
+
+batch.on(:complete, "SomeClass.on_complete", kw_param: 1)
+batch.on(:success, "SomeClass.on_success", some_param: 'foo')
+
+# Add context to the batch. Can be accessed as batch_context on any jobs within the batch.
+# Nested Batches will have their contexts merged
+batch.context = {
+  some_value: 'blah',
+}
+
+batch.jobs do
+  # Enqueue jobs like normal
+end
+```
+
+#### Job Pools
+A job pool is like a custom Sidekiq Queue. You can add jobs to it and it will empty itself out into one of the actual queues.
+However, it adds some options for tweaking the logic:
+- `concurrency` (default: `nil`) - Define how many jobs from the pool can run at once.
+- `order` (default: `fifo`) - Define how the pool will empty itself
+  - `fifo` - First-In First-Out, a traditional queue
+  - `lifo` - Last-In First-Out
+  - `random` - Pluck and run jobs in random order
+  - `priority` - Execute jobs in a priority order (NB: Due to Redis, this priority-random, meaning that items with the same priority will be run in random order, not fifo)
+- `clean_when_empty` (default: `true`) - Automatically clean the pool when it is empty
+- `on_failed_job` (default `:wait`) - If a Job fails, should the pool `:continue` and still enqueue the next job or `:wait` for the job to succeed
+
+Example:
+```ruby
+pool = CanvasSync::JobBatches::Pool.new(concurrency: 4, order: :priority, clean_when_empty: false)
+pool_id = pool.pid
+
+# Add a job to the pool
+pool << {
+  job: SomeJob, # The Class of a ActiveJob Job or Sidekiq Worker
+  parameters: [1, 2, 3], # Array of params to pass th e Job
+  priority: 100, # Only effective if order=:priority, higher is higher
+}
+
+# Add many jobs to the pool
+pool.add_jobs([
+  {
+    job: SomeJob, # The Class of a ActiveJob Job or Sidekiq Worker
+    parameters: [1, 2, 3], # Array of params to pass th e Job
+    priority: 100, # Only effective if order=:priority, higher is higher
+  },
+  # ...
+])
+
+# ...Later
+CanvasSync::JobBatches::Pool.from_pid(pool_id).cleanup_redis
+
+```
+
 ## Legacy Support
 
 ### Legacy Mappings

data/lib/canvas_sync/job_batches/batch.rb

@@ -4,10 +4,13 @@ begin
 rescue LoadError
 end
 
+require_relative './redis_model'
+require_relative './redis_script'
 require_relative './batch_aware_job'
 require_relative "./callback"
 require_relative "./context_hash"
 require_relative "./status"
+require_relative "./pool"
 Dir[File.dirname(__FILE__) + "/jobs/*.rb"].each { |file| require file }
 require_relative "./chain_builder"
 
@@ -17,24 +20,9 @@ require_relative "./chain_builder"
 module CanvasSync
   module JobBatches
     class Batch
-      class NoBlockGivenError < StandardError; end
-
-      def self.batch_attr(key, read_only: true)
-        class_eval <<-RUBY, __FILE__, __LINE__ + 1
-          def #{key}=(value)
-            raise "#{key} is read-only once the batch has been started" if #{read_only.to_s} && (@initialized || @existing)
-            @#{key} = value
-            persist_bid_attr('#{key}', value)
-          end
+      include RedisModel
 
-          def #{key}
-            return @#{key} if defined?(@#{key})
-            if (@initialized || @existing)
-              @#{key} = read_bid_attr('#{key}')
-            end
-          end
-        RUBY
-      end
+      class NoBlockGivenError < StandardError; end
 
       delegate :redis, to: :class
 
@@ -47,16 +35,14 @@ module CanvasSync
         @existing = !(!existing_bid || existing_bid.empty?)  # Basically existing_bid.present?
         @initialized = false
         @bidkey = "BID-" + @bid.to_s
-        @pending_attrs = {}
-        @ready_to_queue = nil
         self.created_at = Time.now.utc.to_f unless @existing
       end
 
-      batch_attr :description
-      batch_attr :created_at
-      batch_attr :callback_queue, read_only: false
-      batch_attr :callback_batch, read_only: false
-      batch_attr :allow_context_changes
+      redis_attr :description
+      redis_attr :created_at
+      redis_attr :callback_queue, read_only: false
+      redis_attr :callback_params, :json
+      redis_attr :allow_context_changes
 
       def context
         return @context if defined?(@context)
@@ -122,29 +108,24 @@ module CanvasSync
           @context&.save!
 
           @initialized = true
+        else
+          assert_batch_is_open
         end
 
-        job_queue = @ready_to_queue = []
-
         begin
           parent = Thread.current[:batch]
           Thread.current[:batch] = self
           yield
         ensure
-          @ready_to_queue = nil
-          append_jobs(job_queue, parent_bid)
           Thread.current[:batch] = parent
         end
 
-        job_queue
+        nil
       end
 
       def increment_job_queue(jid)
-        if @ready_to_queue
-          @ready_to_queue << jid
-        else
-          append_jobs([jid])
-        end
+        assert_batch_is_open
+        append_jobs([jid])
       end
 
       def invalidate_all
@@ -170,52 +151,42 @@ module CanvasSync
         batch.parent ? valid && valid?(batch.parent) : valid
       end
 
-      # Any Batches or Jobs created in the given block won't be assocaiated to the current batch
-      def self.without_batch
+      def self.with_batch(batch)
+        batch = self.new(batch) if batch.is_a?(String)
         parent = Thread.current[:batch]
-        Thread.current[:batch] = nil
+        Thread.current[:batch] = batch
         yield
       ensure
         Thread.current[:batch] = parent
       end
 
-      private
-
-      def persist_bid_attr(attribute, value)
-        if @initialized || @existing
-          redis do |r|
-            r.multi do
-              r.hset(@bidkey, attribute, value)
-              r.expire(@bidkey, BID_EXPIRE_TTL)
-            end
-          end
-        else
-          @pending_attrs[attribute] = value
-        end
+      # Any Batches or Jobs created in the given block won't be assocaiated to the current batch
+      def self.without_batch(&blk)
+        with_batch(nil, &blk)
       end
 
-      def read_bid_attr(attribute)
-        redis do |r|
-          r.hget(@bidkey, attribute)
-        end
+      protected
+
+      def redis_key
+        @bidkey
       end
 
-      def flush_pending_attrs
-        redis do |r|
-          r.mapped_hmset(@bidkey, @pending_attrs)
+      private
+
+      def assert_batch_is_open
+        unless defined?(@closed)
+          redis do |r|
+            @closed = r.hget(@bidkey, 'success') == 'true'
+          end
         end
-        @pending_attrs = {}
+        raise "Cannot add jobs to Batch #{} bid - it has already entered the callback-stage" if @closed
       end
 
-      def append_jobs(jids, parent_bid = self.parent_bid)
+      def append_jobs(jids)
+        jids = jids.uniq
         redis do |r|
           r.multi do
-            if parent_bid
-              r.hincrby("BID-#{parent_bid}", "total", jids.size)
-            end
-
             r.hincrby(@bidkey, "pending", jids.size)
-            r.hincrby(@bidkey, "total", jids.size)
             r.expire(@bidkey, BID_EXPIRE_TTL)
 
             if jids.size > 0
@@ -242,17 +213,6 @@ module CanvasSync
             end
           end
 
-          # if the batch failed, and has a parent, update the parent to show one pending and failed job
-          if parent_bid
-            redis do |r|
-              r.multi do
-                r.hincrby("BID-#{parent_bid}", "pending", 1)
-                r.sadd("BID-#{parent_bid}-failed", jid)
-                r.expire("BID-#{parent_bid}-failed", BID_EXPIRE_TTL)
-              end
-            end
-          end
-
           if pending.to_i == failed.to_i && children == complete
             enqueue_callbacks(:complete, bid)
           end
@@ -285,7 +245,7 @@ module CanvasSync
         end
 
         def process_successful_job(bid, jid)
-          _, failed, pending, children, complete, success, total, parent_bid = redis do |r|
+          _, failed, pending, children, complete, success, parent_bid = redis do |r|
             r.multi do
               r.srem("BID-#{bid}-failed", jid)
 
@@ -294,14 +254,16 @@ module CanvasSync
               r.hincrby("BID-#{bid}", "children", 0)
               r.scard("BID-#{bid}-batches-complete")
               r.scard("BID-#{bid}-batches-success")
-              r.hget("BID-#{bid}", "total")
               r.hget("BID-#{bid}", "parent_bid")
 
+              r.hincrby("BID-#{bid}", "successful-jobs", 1)
               r.srem("BID-#{bid}-jids", jid)
               r.expire("BID-#{bid}", BID_EXPIRE_TTL)
             end
           end
 
+          # TODO - There seems to be an issue where a :complete callback batch will occasionally run its job, but won't enqueue_callbacks()
+
           all_success = pending.to_i.zero? && children == success
           # if complete or successfull call complete callback (the complete callback may then call successful)
           if (pending.to_i == failed.to_i && children == complete) || all_success
@@ -313,14 +275,15 @@ module CanvasSync
         def enqueue_callbacks(event, bid)
           batch_key = "BID-#{bid}"
           callback_key = "#{batch_key}-callbacks-#{event}"
-          already_processed, _, callbacks, queue, parent_bid, callback_batch = redis do |r|
+          already_processed, _, callbacks, queue, parent_bid, callback_params = redis do |r|
+            return unless r.exists?(batch_key)
             r.multi do
               r.hget(batch_key, event)
               r.hset(batch_key, event, true)
               r.smembers(callback_key)
               r.hget(batch_key, "callback_queue")
               r.hget(batch_key, "parent_bid")
-              r.hget(batch_key, "callback_batch")
+              r.hget(batch_key, "callback_params")
             end
           end
 
@@ -328,6 +291,7 @@ module CanvasSync
 
           queue ||= "default"
           parent_bid = !parent_bid || parent_bid.empty? ? nil : parent_bid    # Basically parent_bid.blank?
+          callback_params = JSON.parse(callback_params) if callback_params.present?
           callback_args = callbacks.reduce([]) do |memo, jcb|
             cb = JSON.load(jcb)
             memo << [cb['callback'], event.to_s, cb['opts'], bid, parent_bid]
@@ -335,38 +299,32 @@ module CanvasSync
 
           opts = {"bid" => bid, "event" => event}
 
-          # Run callback batch finalize synchronously
-          if callback_batch
-            # Extract opts from cb_args or use current
-            # Pass in stored event as callback finalize is processed on complete event
-            cb_opts = callback_args.first&.at(2) || opts
+          if callback_args.present? && !callback_params.present?
+            logger.debug {"Enqueue callback bid: #{bid} event: #{event} args: #{callback_args.inspect}"}
 
-            logger.debug {"Run callback batch bid: #{bid} event: #{event} args: #{callback_args.inspect}"}
-            # Finalize now
-            finalizer = Batch::Callback::Finalize.new
-            status = Status.new bid
-            finalizer.dispatch(status, cb_opts)
+            with_batch(parent_bid) do
+              cb_batch = self.new
+              cb_batch.callback_params = {
+                for_bid: bid,
+                event: event,
+              }
+              opts['callback_bid'] = cb_batch.bid
 
-            return
+              logger.debug {"Adding callback batch: #{cb_batch.bid} for batch: #{bid}"}
+              cb_batch.jobs do
+                push_callbacks callback_args, queue
+              end
+            end
           end
 
-          logger.debug {"Enqueue callback bid: #{bid} event: #{event} args: #{callback_args.inspect}"}
-
-          if callback_args.empty?
-            # Finalize now
-            finalizer = Batch::Callback::Finalize.new
-            status = Status.new bid
-            finalizer.dispatch(status, opts)
-          else
-            # Otherwise finalize in sub batch complete callback
-            cb_batch = self.new
-            cb_batch.callback_batch = true
-            logger.debug {"Adding callback batch: #{cb_batch.bid} for batch: #{bid}"}
-            cb_batch.on(:complete, "#{Batch::Callback::Finalize.to_s}#dispatch", opts)
-            cb_batch.jobs do
-              push_callbacks callback_args, queue
-            end
+          if callback_params.present?
+            opts['origin'] = callback_params
           end
+
+          logger.debug {"Run batch finalizer bid: #{bid} event: #{event} args: #{callback_args.inspect}"}
+          finalizer = Batch::Callback::Finalize.new
+          status = Status.new bid
+          finalizer.dispatch(status, opts)
         end
 
         def cleanup_redis(bid)
@@ -381,6 +339,7 @@ module CanvasSync
               "BID-#{bid}-batches-success",
               "BID-#{bid}-batches-complete",
               "BID-#{bid}-batches-failed",
+              "BID-#{bid}-bids",
               "BID-#{bid}-jids",
             )
           end
@@ -397,7 +356,7 @@ module CanvasSync
         private
 
         def push_callbacks(args, queue)
-          Batch::Callback::Worker.enqueue_all(args, queue)
+          Batch::Callback::worker_class.enqueue_all(args, queue)
         end
       end
     end

data/lib/canvas_sync/job_batches/callback.rb

@@ -2,6 +2,7 @@ module CanvasSync
   module JobBatches
     class Batch
       module Callback
+        mattr_accessor :worker_class
 
         VALID_CALLBACKS = %w[success complete dead].freeze
 
@@ -47,47 +48,34 @@ module CanvasSync
           end
         end
 
-        if defined?(::Sidekiq)
-          class SidekiqCallbackWorker
-            include ::Sidekiq::Worker
-            include CallbackWorkerCommon
-
-            def self.enqueue_all(args, queue)
-              return if args.empty?
-
-              ::Sidekiq::Client.push_bulk(
-                'class' => self,
-                'args' => args,
-                'queue' => queue
-              )
-            end
-          end
-          Worker = SidekiqCallbackWorker
-        else
-          Worker = ActiveJobCallbackWorker
-        end
+        worker_class = ActiveJobCallbackWorker
 
         class Finalize
-          def dispatch status, opts
+          def dispatch(status, opts)
+            is_callback_batch = opts['origin'].present?
+            has_callback_batch = opts['callback_bid'].present?
+
             bid = opts["bid"]
-            callback_bid = status.bid
             event = opts["event"].to_sym
-            callback_batch = bid != callback_bid
 
-            Batch.logger.debug {"Finalize #{event} batch id: #{opts["bid"]}, callback batch id: #{callback_bid} callback_batch #{callback_batch}"}
+            Batch.logger.debug {"Finalize #{event} batch id: #{opts["bid"]}, callback batch id: #{callback_bid} callback_batch #{is_callback_batch}"}
 
             batch_status = Status.new bid
             send(event, bid, batch_status, batch_status.parent_bid)
 
-            # Different events are run in different callback batches
-            Batch.cleanup_redis callback_bid if callback_batch
-            Batch.cleanup_redis bid if event == :success
+            if event == :success && !has_callback_batch
+              Batch.cleanup_redis(bid)
+            end
+
+            if event == :success && is_callback_batch && opts['origin']['event'].to_sym == :success
+              Batch.cleanup_redis(opts['origin']['for_bid'])
+            end
           end
 
           def success(bid, status, parent_bid)
             return unless parent_bid
 
-            _, _, success, _, _, complete, pending, children, failure = Batch.redis do |r|
+            _, _, success, _, _, complete, pending, children, success, failure = Batch.redis do |r|
               r.multi do
                 r.sadd("BID-#{parent_bid}-batches-success", bid)
                 r.expire("BID-#{parent_bid}-batches-success", Batch::BID_EXPIRE_TTL)
@@ -99,15 +87,21 @@ module CanvasSync
 
                 r.hincrby("BID-#{parent_bid}", "pending", 0)
                 r.hincrby("BID-#{parent_bid}", "children", 0)
+                r.scard("BID-#{parent_bid}-batches-success")
                 r.scard("BID-#{parent_bid}-failed")
               end
             end
-            # if job finished successfully and parent batch completed call parent complete callback
-            # Success callback is called after complete callback
+
+            # If the job finished successfully and parent batch is completed, call parent :complete callback
+            # Parent :success callback will be called by its :complete callback
             if complete == children && pending == failure
               Batch.logger.debug {"Finalize parent complete bid: #{parent_bid}"}
               Batch.enqueue_callbacks(:complete, parent_bid)
             end
+            if pending.to_i.zero? && children == success
+              Batch.logger.debug {"Finalize parent success bid: #{parent_bid}"}
+              Batch.enqueue_callbacks(:success, parent_bid)
+            end
           end
 
           def complete(bid, status, parent_bid)
@@ -119,10 +113,12 @@ module CanvasSync
               end
             end
 
-            # if we batch was successful run success callback
+            # If the batch was successful run :success callback, which will call the parent's :complete callback (if necessary)
+            # Also, only trigger the success callback if the :complete callback_batch was successful
             if pending.to_i.zero? && children == success
-              Batch.enqueue_callbacks(:success, bid)
+              # Batch.enqueue_callbacks(:success, bid)
 
+            # otherwise check for a parent and call its :complete if needed
             elsif parent_bid
               # if batch was not successfull check and see if its parent is complete
               # if the parent is complete we trigger the complete callback