pgbus 0.3.6 → 0.3.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bd4586e94a76eaf20987039101a038af4e64a9d3bcfd5d68b5b2ef984f3459f3
-  data.tar.gz: 98f2e88891c3e7f00963d697d4fc135271a61eccef04aa89dbfb40e0d568b240
+  metadata.gz: 76f2d5b122b3a379c36fcef393e8b7eabcfdbc7c8534d76948b1349dd97cb58b
+  data.tar.gz: 4059948c68407aebeebd6f45d3b550158c6d38455c95cf255ccf1faf2d10889a
 SHA512:
-  metadata.gz: f2df68ef64f8b4ac37dfe509c800688c51e811910855f69807c28a5e3777d0eff10dc989ea4e0180b17ee9feada2549a6faa182052c2bbf88be225ccfa7a45d2
-  data.tar.gz: 4890d5ab0a55df38ba8d74bb704c00a6d82a4abe17784600e19615288b6dfebec50ac07ac7a89a0cdffa95c85299b795071d4a5dbcc9d351bac076a5baca8bc9
+  metadata.gz: b21741c2de21716ce47c0246c8de08bb440e9a4dfc750bc60e95484b97fe9c7626b357c47c4b2b7583c6d8273fef08b363aa84ed5424bf21a64bb64519ce64b7
+  data.tar.gz: 48983a1abcbfc000936caf9ff2cac0d7451633aee2ee05c68c4168b84c0aed4ee38e4108501214b91c5045e9cc97f269f55c3e9e6050d778ac06b07a11c6c45d

data/lib/generators/pgbus/add_failed_events_index_generator.rb

@@ -0,0 +1,51 @@
+# frozen_string_literal: true
+
+require "rails/generators"
+require "rails/generators/active_record"
+
+module Pgbus
+  module Generators
+    class AddFailedEventsIndexGenerator < Rails::Generators::Base
+      include ActiveRecord::Generators::Migration
+
+      source_root File.expand_path("templates", __dir__)
+
+      desc "Add unique index on pgbus_failed_events (queue_name, msg_id) for failure tracking upserts"
+
+      class_option :database,
+                   type: :string,
+                   default: nil,
+                   desc: "Use a separate database for pgbus tables (e.g. --database=pgbus)"
+
+      def create_migration_file
+        if separate_database?
+          migration_template "add_failed_events_unique_index.rb.erb",
+                             "db/pgbus_migrate/add_pgbus_failed_events_unique_index.rb"
+        else
+          migration_template "add_failed_events_unique_index.rb.erb",
+                             "db/migrate/add_pgbus_failed_events_unique_index.rb"
+        end
+      end
+
+      def display_post_install
+        say ""
+        say "Pgbus failed events unique index added!", :green
+        say ""
+        say "Next steps:"
+        say "  1. Run: rails db:migrate#{":#{options[:database]}" if separate_database?}"
+        say "  2. Failed jobs will now be tracked in the dashboard"
+        say ""
+      end
+
+      private
+
+      def migration_version
+        "[#{ActiveRecord::Migration.current_version}]"
+      end
+
+      def separate_database?
+        options[:database].present?
+      end
+    end
+  end
+end

data/lib/generators/pgbus/templates/add_failed_events_unique_index.rb.erb

@@ -0,0 +1,7 @@
+class AddPgbusFailedEventsUniqueIndex < ActiveRecord::Migration<%= migration_version %>
+  def change
+    add_index :pgbus_failed_events, [:queue_name, :msg_id],
+      unique: true, name: "idx_pgbus_failed_events_queue_msg",
+      if_not_exists: true
+  end
+end

data/lib/generators/pgbus/templates/migration.rb.erb

@@ -65,6 +65,8 @@ class CreatePgbusTables < ActiveRecord::Migration<%= migration_version %>
 
     add_index :pgbus_failed_events, :queue_name, name: "idx_pgbus_failed_events_queue"
     add_index :pgbus_failed_events, :failed_at, name: "idx_pgbus_failed_events_time"
+    add_index :pgbus_failed_events, [:queue_name, :msg_id],
+      unique: true, name: "idx_pgbus_failed_events_queue_msg"
 
     # Concurrency semaphores (counting locks for job concurrency limits)
     create_table :pgbus_semaphores do |t|

data/lib/pgbus/active_job/executor.rb

@@ -20,6 +20,7 @@ module Pgbus
 
         if read_count > config.max_retries
           handle_dead_letter(message, queue_name, payload, source_queue: source_queue)
+          FailedEventRecorder.clear!(queue_name: queue_name, msg_id: message.msg_id.to_i)
           signal_concurrency(payload)
           signal_batch_discarded(payload)
           Uniqueness.release_lock(Uniqueness.extract_key(payload))
@@ -54,6 +55,7 @@ module Pgbus
           job = ::ActiveJob::Base.deserialize(payload)
           execute_job(job)
           archive_from(queue_name, message.msg_id.to_i, source_queue: source_queue)
+          FailedEventRecorder.clear!(queue_name: queue_name, msg_id: message.msg_id.to_i)
           job_succeeded = true
         end
 
@@ -61,7 +63,7 @@ module Pgbus
         record_stat(payload, queue_name, "success", execution_start, message: message)
         :success
       rescue StandardError => e
-        handle_failure(message, queue_name, e)
+        handle_failure(message, queue_name, e, payload: payload)
         instrument("pgbus.job_failed", queue: queue_name, job_class: payload&.dig("job_class"), error: e.class.name)
         record_stat(payload, queue_name, "failed", execution_start, message: message)
         # Don't signal concurrency on transient failure — the job will be retried.
@@ -143,13 +145,22 @@ module Pgbus
         [((Time.now.utc - enqueued_at) * 1000).round, 0].max
       end
 
-      def handle_failure(_message, _queue_name, error)
+      def handle_failure(message, queue_name, error, payload: nil)
         Pgbus.logger.error { "[Pgbus] Job failed: #{error.class}: #{error.message}" }
         Pgbus.logger.debug { error.backtrace&.join("\n") }
 
+        # Record failure for dashboard visibility.
         # Message visibility timeout will expire and it becomes available again.
         # read_ct tracks delivery attempts — when it exceeds max_retries,
         # the next read will route to DLQ.
+        FailedEventRecorder.record!(
+          queue_name: queue_name,
+          msg_id: message.msg_id.to_i,
+          payload: payload || message.message,
+          headers: message.respond_to?(:headers) ? message.headers : nil,
+          error: error,
+          retry_count: [message.read_ct.to_i - 1, 0].max
+        )
       end
 
       def instrument(event_name, payload = {})

data/lib/pgbus/failed_event_recorder.rb

@@ -0,0 +1,59 @@
+# frozen_string_literal: true
+
+module Pgbus
+  # Records job failures to pgbus_failed_events for dashboard visibility.
+  # Uses upsert (INSERT ON CONFLICT UPDATE) keyed on (queue_name, msg_id)
+  # so each message has at most one failed_event row tracking its latest error.
+  class FailedEventRecorder
+    class << self
+      def record!(queue_name:, msg_id:, payload:, headers:, error:, retry_count:)
+        connection.exec_query(
+          <<~SQL.squish,
+            INSERT INTO pgbus_failed_events
+              (queue_name, msg_id, payload, headers, error_class, error_message, backtrace, retry_count, failed_at)
+            VALUES ($1, $2, $3, $4, $5, $6, $7, $8, CURRENT_TIMESTAMP)
+            ON CONFLICT (queue_name, msg_id) DO UPDATE SET
+              error_class = EXCLUDED.error_class,
+              error_message = EXCLUDED.error_message,
+              backtrace = EXCLUDED.backtrace,
+              retry_count = EXCLUDED.retry_count,
+              failed_at = EXCLUDED.failed_at
+          SQL
+          "FailedEvent Record",
+          [
+            queue_name,
+            msg_id.to_i,
+            payload.is_a?(String) ? payload : JSON.generate(payload),
+            headers.is_a?(String) ? headers : headers&.then { |h| JSON.generate(h) },
+            error.class.name,
+            error.message.to_s.truncate(10_000),
+            error.backtrace&.first(30)&.join("\n"),
+            retry_count
+          ]
+        )
+      rescue StandardError => e
+        Pgbus.logger.debug { "[Pgbus] Failed to record failed event: #{e.message}" }
+      end
+
+      def clear!(queue_name:, msg_id:)
+        connection.exec_delete(
+          "DELETE FROM pgbus_failed_events WHERE queue_name = $1 AND msg_id = $2",
+          "FailedEvent Clear",
+          [queue_name, msg_id.to_i]
+        )
+      rescue StandardError => e
+        Pgbus.logger.debug { "[Pgbus] Failed to clear failed event: #{e.message}" }
+      end
+
+      private
+
+      def connection
+        if defined?(BusRecord) && BusRecord.connected?
+          BusRecord.connection
+        else
+          ActiveRecord::Base.connection
+        end
+      end
+    end
+  end
+end

data/lib/pgbus/recurring/schedule.rb

@@ -151,8 +151,11 @@ module Pgbus
           :already_locked
         end
       rescue StandardError => e
-        Pgbus.logger.warn { "[Pgbus] Uniqueness lock failed for #{task.key}: #{e.message}" }
-        nil # Fail open — allow enqueue if lock check errors
+        Pgbus.logger.error do
+          "[Pgbus] Uniqueness lock check failed for #{task.key}: #{e.class}: #{e.message} — " \
+            "skipping enqueue to prevent duplicates"
+        end
+        :already_locked # Fail closed — skip enqueue when lock check errors
       end
 
       # Release a uniqueness lock. Safe to call with nil or :already_locked.

data/lib/pgbus/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Pgbus
-  VERSION = "0.3.6"
+  VERSION = "0.3.8"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pgbus
 version: !ruby/object:Gem::Version
-  version: 0.3.6
+  version: 0.3.8
 platform: ruby
 authors:
 - Mikael Henriksson
@@ -198,6 +198,7 @@ files:
 - config/routes.rb
 - exe/pgbus
 - lib/active_job/queue_adapters/pgbus_adapter.rb
+- lib/generators/pgbus/add_failed_events_index_generator.rb
 - lib/generators/pgbus/add_job_locks_generator.rb
 - lib/generators/pgbus/add_job_stats_generator.rb
 - lib/generators/pgbus/add_job_stats_latency_generator.rb
@@ -206,6 +207,7 @@ files:
 - lib/generators/pgbus/add_recurring_generator.rb
 - lib/generators/pgbus/install_generator.rb
 - lib/generators/pgbus/migrate_job_locks_generator.rb
+- lib/generators/pgbus/templates/add_failed_events_unique_index.rb.erb
 - lib/generators/pgbus/templates/add_job_locks.rb.erb
 - lib/generators/pgbus/templates/add_job_stats.rb.erb
 - lib/generators/pgbus/templates/add_job_stats_latency.rb.erb
@@ -240,6 +242,7 @@ files:
 - lib/pgbus/event_bus/publisher.rb
 - lib/pgbus/event_bus/registry.rb
 - lib/pgbus/event_bus/subscriber.rb
+- lib/pgbus/failed_event_recorder.rb
 - lib/pgbus/instrumentation.rb
 - lib/pgbus/outbox.rb
 - lib/pgbus/outbox/poller.rb