sidekiq 6.0.0 → 6.4.0

data/lib/sidekiq/api.rb

@@ -2,25 +2,13 @@
 
 require "sidekiq"
 
-module Sidekiq
-  module RedisScanner
-    def sscan(conn, key)
-      cursor = "0"
-      result = []
-      loop do
-        cursor, values = conn.sscan(key, cursor)
-        result.push(*values)
-        break if cursor == "0"
-      end
-      result
-    end
-  end
+require "zlib"
+require "base64"
 
+module Sidekiq
   class Stats
-    include RedisScanner
-
     def initialize
-      fetch_stats!
+      fetch_stats_fast!
     end
 
     def processed
@@ -63,7 +51,8 @@ module Sidekiq
       Sidekiq::Stats::Queues.new.lengths
     end
 
-    def fetch_stats!
+    # O(1) redis calls
+    def fetch_stats_fast!
       pipe1_res = Sidekiq.redis { |conn|
         conn.pipelined do
           conn.get("stat:processed")
@@ -76,37 +65,19 @@ module Sidekiq
         end
       }
 
-      processes = Sidekiq.redis { |conn|
-        sscan(conn, "processes")
-      }
-
-      queues = Sidekiq.redis { |conn|
-        sscan(conn, "queues")
-      }
-
-      pipe2_res = Sidekiq.redis { |conn|
-        conn.pipelined do
-          processes.each { |key| conn.hget(key, "busy") }
-          queues.each { |queue| conn.llen("queue:#{queue}") }
-        end
-      }
-
-      s = processes.size
-      workers_size = pipe2_res[0...s].map(&:to_i).inject(0, &:+)
-      enqueued = pipe2_res[s..-1].map(&:to_i).inject(0, &:+)
-
       default_queue_latency = if (entry = pipe1_res[6].first)
         job = begin
-                Sidekiq.load_json(entry)
-              rescue
-                {}
-              end
+          Sidekiq.load_json(entry)
+        rescue
+          {}
+        end
         now = Time.now.to_f
         thence = job["enqueued_at"] || now
         now - thence
       else
         0
       end
+
       @stats = {
         processed: pipe1_res[0].to_i,
         failed: pipe1_res[1].to_i,
@@ -115,10 +86,39 @@ module Sidekiq
         dead_size: pipe1_res[4],
         processes_size: pipe1_res[5],
 
-        default_queue_latency: default_queue_latency,
-        workers_size: workers_size,
-        enqueued: enqueued,
+        default_queue_latency: default_queue_latency
+      }
+    end
+
+    # O(number of processes + number of queues) redis calls
+    def fetch_stats_slow!
+      processes = Sidekiq.redis { |conn|
+        conn.sscan_each("processes").to_a
+      }
+
+      queues = Sidekiq.redis { |conn|
+        conn.sscan_each("queues").to_a
+      }
+
+      pipe2_res = Sidekiq.redis { |conn|
+        conn.pipelined do
+          processes.each { |key| conn.hget(key, "busy") }
+          queues.each { |queue| conn.llen("queue:#{queue}") }
+        end
       }
+
+      s = processes.size
+      workers_size = pipe2_res[0...s].sum(&:to_i)
+      enqueued = pipe2_res[s..-1].sum(&:to_i)
+
+      @stats[:workers_size] = workers_size
+      @stats[:enqueued] = enqueued
+      @stats
+    end
+
+    def fetch_stats!
+      fetch_stats_fast!
+      fetch_stats_slow!
     end
 
     def reset(*stats)
@@ -138,15 +138,14 @@ module Sidekiq
     private
 
     def stat(s)
-      @stats[s]
+      fetch_stats_slow! if @stats[s].nil?
+      @stats[s] || raise(ArgumentError, "Unknown stat #{s}")
     end
 
     class Queues
-      include RedisScanner
-
       def lengths
         Sidekiq.redis do |conn|
-          queues = sscan(conn, "queues")
+          queues = conn.sscan_each("queues").to_a
 
           lengths = conn.pipelined {
             queues.each do |queue|
@@ -154,19 +153,16 @@ module Sidekiq
             end
           }
 
-          i = 0
-          array_of_arrays = queues.each_with_object({}) { |queue, memo|
-            memo[queue] = lengths[i]
-            i += 1
-          }.sort_by { |_, size| size }
-
-          Hash[array_of_arrays.reverse]
+          array_of_arrays = queues.zip(lengths).sort_by { |_, size| -size }
+          array_of_arrays.to_h
         end
       end
     end
 
     class History
       def initialize(days_previous, start_date = nil)
+        # we only store five years of data in Redis
+        raise ArgumentError if days_previous < 1 || days_previous > (5 * 365)
         @days_previous = days_previous
         @start_date = start_date || Time.now.utc.to_date
       end
@@ -182,18 +178,12 @@ module Sidekiq
       private
 
       def date_stat_hash(stat)
-        i = 0
         stat_hash = {}
-        keys = []
-        dates = []
-
-        while i < @days_previous
-          date = @start_date - i
-          datestr = date.strftime("%Y-%m-%d")
-          keys << "stat:#{stat}:#{datestr}"
-          dates << datestr
-          i += 1
-        end
+        dates = @start_date.downto(@start_date - @days_previous + 1).map { |date|
+          date.strftime("%Y-%m-%d")
+        }
+
+        keys = dates.map { |datestr| "stat:#{stat}:#{datestr}" }
 
         begin
           Sidekiq.redis do |conn|
@@ -225,13 +215,12 @@ module Sidekiq
   #
   class Queue
     include Enumerable
-    extend RedisScanner
 
     ##
     # Return all known queues within Redis.
     #
     def self.all
-      Sidekiq.redis { |c| sscan(c, "queues") }.sort.map { |q| Sidekiq::Queue.new(q) }
+      Sidekiq.redis { |c| c.sscan_each("queues").to_a }.sort.map { |q| Sidekiq::Queue.new(q) }
     end
 
     attr_reader :name
@@ -281,7 +270,7 @@ module Sidekiq
         break if entries.empty?
         page += 1
         entries.each do |entry|
-          yield Job.new(entry, @name)
+          yield JobRecord.new(entry, @name)
         end
         deleted_size = initial_size - size
       end
@@ -291,7 +280,7 @@ module Sidekiq
     # Find the job with the given JID within this queue.
     #
     # This is a slow, inefficient operation.  Do not use under
-    # normal conditions.  Sidekiq Pro contains a faster version.
+    # normal conditions.
     def find_job(jid)
       detect { |j| j.jid == jid }
     end
@@ -299,7 +288,7 @@ module Sidekiq
     def clear
       Sidekiq.redis do |conn|
         conn.multi do
-          conn.del(@rname)
+          conn.unlink(@rname)
           conn.srem("queues", name)
         end
       end
@@ -312,9 +301,9 @@ module Sidekiq
   # sorted set.
   #
   # The job should be considered immutable but may be
-  # removed from the queue via Job#delete.
+  # removed from the queue via JobRecord#delete.
   #
-  class Job
+  class JobRecord
     attr_reader :item
     attr_reader :value
 
@@ -342,21 +331,23 @@ module Sidekiq
 
     def display_class
       # Unwrap known wrappers so they show up in a human-friendly manner in the Web UI
-      @klass ||= case klass
-                 when /\ASidekiq::Extensions::Delayed/
-                   safe_load(args[0], klass) do |target, method, _|
-                     "#{target}.#{method}"
-                   end
-                 when "ActiveJob::QueueAdapters::SidekiqAdapter::JobWrapper"
-                   job_class = @item["wrapped"] || args[0]
-                   if job_class == "ActionMailer::DeliveryJob"
-                     # MailerClass#mailer_method
-                     args[0]["arguments"][0..1].join("#")
-                   else
-                     job_class
-                   end
-                 else
-                   klass
+      @klass ||= self["display_class"] || begin
+        case klass
+        when /\ASidekiq::Extensions::Delayed/
+          safe_load(args[0], klass) do |target, method, _|
+            "#{target}.#{method}"
+          end
+        when "ActiveJob::QueueAdapters::SidekiqAdapter::JobWrapper"
+          job_class = @item["wrapped"] || args[0]
+          if job_class == "ActionMailer::DeliveryJob" || job_class == "ActionMailer::MailDeliveryJob"
+            # MailerClass#mailer_method
+            args[0]["arguments"][0..1].join("#")
+          else
+            job_class
+          end
+        else
+          klass
+        end
       end
     end
 
@@ -372,6 +363,9 @@ module Sidekiq
                   if (self["wrapped"] || args[0]) == "ActionMailer::DeliveryJob"
                     # remove MailerClass, mailer_method and 'deliver_now'
                     job_args.drop(3)
+                  elsif (self["wrapped"] || args[0]) == "ActionMailer::MailDeliveryJob"
+                    # remove MailerClass, mailer_method and 'deliver_now'
+                    job_args.drop(3).first["args"]
                   else
                     job_args
                   end
@@ -400,6 +394,20 @@ module Sidekiq
       Time.at(self["created_at"] || self["enqueued_at"] || 0).utc
     end
 
+    def tags
+      self["tags"] || []
+    end
+
+    def error_backtrace
+      # Cache nil values
+      if defined?(@error_backtrace)
+        @error_backtrace
+      else
+        value = self["error_backtrace"]
+        @error_backtrace = value && uncompress_backtrace(value)
+      end
+    end
+
     attr_reader :queue
 
     def latency
@@ -433,9 +441,26 @@ module Sidekiq
       Sidekiq.logger.warn "Unable to load YAML: #{ex.message}" unless Sidekiq.options[:environment] == "development"
       default
     end
+
+    def uncompress_backtrace(backtrace)
+      if backtrace.is_a?(Array)
+        # Handle old jobs with raw Array backtrace format
+        backtrace
+      else
+        decoded = Base64.decode64(backtrace)
+        uncompressed = Zlib::Inflate.inflate(decoded)
+        begin
+          Sidekiq.load_json(uncompressed)
+        rescue
+          # Handle old jobs with marshalled backtrace format
+          # TODO Remove in 7.x
+          Marshal.load(uncompressed)
+        end
+      end
+    end
   end
 
-  class SortedEntry < Job
+  class SortedEntry < JobRecord
     attr_reader :score
     attr_reader :parent
 
@@ -458,8 +483,9 @@ module Sidekiq
     end
 
     def reschedule(at)
-      delete
-      @parent.schedule(at, item)
+      Sidekiq.redis do |conn|
+        conn.zincrby(@parent.name, at.to_f - @score, Sidekiq.dump_json(@item))
+      end
     end
 
     def add_to_queue
@@ -503,7 +529,7 @@ module Sidekiq
         else
           # multiple jobs with the same score
           # find the one with the right JID and push it
-          hash = results.group_by { |message|
+          matched, nonmatched = results.partition { |message|
             if message.index(jid)
               msg = Sidekiq.load_json(message)
               msg["jid"] == jid
@@ -512,12 +538,12 @@ module Sidekiq
             end
           }
 
-          msg = hash.fetch(true, []).first
+          msg = matched.first
           yield msg if msg
 
           # push the rest back onto the sorted set
           conn.multi do
-            hash.fetch(false, []).each do |message|
+            nonmatched.each do |message|
               conn.zadd(parent.name, score.to_f.to_s, message)
             end
           end
@@ -540,9 +566,20 @@ module Sidekiq
       Sidekiq.redis { |c| c.zcard(name) }
     end
 
+    def scan(match, count = 100)
+      return to_enum(:scan, match, count) unless block_given?
+
+      match = "*#{match}*" unless match.include?("*")
+      Sidekiq.redis do |conn|
+        conn.zscan_each(name, match: match, count: count) do |entry, score|
+          yield SortedEntry.new(self, score, entry)
+        end
+      end
+    end
+
     def clear
       Sidekiq.redis do |conn|
-        conn.del(name)
+        conn.unlink(name)
       end
     end
     alias_method :💣, :clear
@@ -576,28 +613,40 @@ module Sidekiq
       end
     end
 
+    ##
+    # Fetch jobs that match a given time or Range. Job ID is an
+    # optional second argument.
     def fetch(score, jid = nil)
+      begin_score, end_score =
+        if score.is_a?(Range)
+          [score.first, score.last]
+        else
+          [score, score]
+        end
+
       elements = Sidekiq.redis { |conn|
-        conn.zrangebyscore(name, score, score)
+        conn.zrangebyscore(name, begin_score, end_score, with_scores: true)
       }
 
       elements.each_with_object([]) do |element, result|
-        entry = SortedEntry.new(self, score, element)
-        if jid
-          result << entry if entry.jid == jid
-        else
-          result << entry
-        end
+        data, job_score = element
+        entry = SortedEntry.new(self, job_score, data)
+        result << entry if jid.nil? || entry.jid == jid
       end
     end
 
     ##
     # Find the job with the given JID within this sorted set.
-    #
-    # This is a slow, inefficient operation.  Do not use under
-    # normal conditions.  Sidekiq Pro contains a faster version.
+    # This is a slower O(n) operation.  Do not use for app logic.
     def find_job(jid)
-      detect { |j| j.jid == jid }
+      Sidekiq.redis do |conn|
+        conn.zscan_each(name, match: "*#{jid}*", count: 100) do |entry, score|
+          job = JSON.parse(entry)
+          matched = job["jid"] == jid
+          return SortedEntry.new(self, score, entry) if matched
+        end
+      end
+      nil
     end
 
     def delete_by_value(name, value)
@@ -612,11 +661,13 @@ module Sidekiq
       Sidekiq.redis do |conn|
         elements = conn.zrangebyscore(name, score, score)
         elements.each do |element|
-          message = Sidekiq.load_json(element)
-          if message["jid"] == jid
-            ret = conn.zrem(name, element)
-            @_size -= 1 if ret
-            break ret
+          if element.index(jid)
+            message = Sidekiq.load_json(element)
+            if message["jid"] == jid
+              ret = conn.zrem(name, element)
+              @_size -= 1 if ret
+              break ret
+            end
           end
         end
       end
@@ -720,7 +771,6 @@ module Sidekiq
   #
   class ProcessSet
     include Enumerable
-    include RedisScanner
 
     def initialize(clean_plz = true)
       cleanup if clean_plz
@@ -731,7 +781,7 @@ module Sidekiq
     def cleanup
       count = 0
       Sidekiq.redis do |conn|
-        procs = sscan(conn, "processes").sort
+        procs = conn.sscan_each("processes").to_a.sort
         heartbeats = conn.pipelined {
           procs.each do |key|
             conn.hget(key, "info")
@@ -741,40 +791,41 @@ module Sidekiq
         # the hash named key has an expiry of 60 seconds.
         # if it's not found, that means the process has not reported
         # in to Redis and probably died.
-        to_prune = []
-        heartbeats.each_with_index do |beat, i|
-          to_prune << procs[i] if beat.nil?
-        end
+        to_prune = procs.select.with_index { |proc, i|
+          heartbeats[i].nil?
+        }
         count = conn.srem("processes", to_prune) unless to_prune.empty?
       end
       count
     end
 
     def each
-      procs = Sidekiq.redis { |conn| sscan(conn, "processes") }.sort
+      result = Sidekiq.redis { |conn|
+        procs = conn.sscan_each("processes").to_a.sort
 
-      Sidekiq.redis do |conn|
         # We're making a tradeoff here between consuming more memory instead of
         # making more roundtrips to Redis, but if you have hundreds or thousands of workers,
         # you'll be happier this way
-        result = conn.pipelined {
+        conn.pipelined do
           procs.each do |key|
-            conn.hmget(key, "info", "busy", "beat", "quiet")
+            conn.hmget(key, "info", "busy", "beat", "quiet", "rss", "rtt_us")
           end
-        }
+        end
+      }
 
-        result.each do |info, busy, at_s, quiet|
-          # If a process is stopped between when we query Redis for `procs` and
-          # when we query for `result`, we will have an item in `result` that is
-          # composed of `nil` values.
-          next if info.nil?
+      result.each do |info, busy, at_s, quiet, rss, rtt|
+        # If a process is stopped between when we query Redis for `procs` and
+        # when we query for `result`, we will have an item in `result` that is
+        # composed of `nil` values.
+        next if info.nil?
 
-          hash = Sidekiq.load_json(info)
-          yield Process.new(hash.merge("busy" => busy.to_i, "beat" => at_s.to_f, "quiet" => quiet))
-        end
+        hash = Sidekiq.load_json(info)
+        yield Process.new(hash.merge("busy" => busy.to_i,
+          "beat" => at_s.to_f,
+          "quiet" => quiet,
+          "rss" => rss.to_i,
+          "rtt_us" => rtt.to_i))
       end
-
-      nil
     end
 
     # This method is not guaranteed accurate since it does not prune the set
@@ -785,6 +836,18 @@ module Sidekiq
       Sidekiq.redis { |conn| conn.scard("processes") }
     end
 
+    # Total number of threads available to execute jobs.
+    # For Sidekiq Enterprise customers this number (in production) must be
+    # less than or equal to your licensed concurrency.
+    def total_concurrency
+      sum { |x| x["concurrency"].to_i }
+    end
+
+    def total_rss_in_kb
+      sum { |x| x["rss"].to_i }
+    end
+    alias_method :total_rss, :total_rss_in_kb
+
     # Returns the identity of the current cluster leader or "" if no leader.
     # This is a Sidekiq Enterprise feature, will always return "" in Sidekiq
     # or Sidekiq Pro.
@@ -834,6 +897,10 @@ module Sidekiq
       self["identity"]
     end
 
+    def queues
+      self["queues"]
+    end
+
     def quiet!
       signal("TSTP")
     end
@@ -864,8 +931,8 @@ module Sidekiq
   end
 
   ##
-  # A worker is a thread that is currently processing a job.
-  # Programmatic access to the current active worker set.
+  # The WorkSet stores the work being done by this Sidekiq cluster.
+  # It tracks the process and thread working on each job.
   #
   # WARNING WARNING WARNING
   #
@@ -873,34 +940,40 @@ module Sidekiq
   # If you call #size => 5 and then expect #each to be
   # called 5 times, you're going to have a bad time.
   #
-  #    workers = Sidekiq::Workers.new
-  #    workers.size => 2
-  #    workers.each do |process_id, thread_id, work|
+  #    works = Sidekiq::WorkSet.new
+  #    works.size => 2
+  #    works.each do |process_id, thread_id, work|
   #      # process_id is a unique identifier per Sidekiq process
   #      # thread_id is a unique identifier per thread
   #      # work is a Hash which looks like:
-  #      # { 'queue' => name, 'run_at' => timestamp, 'payload' => msg }
+  #      # { 'queue' => name, 'run_at' => timestamp, 'payload' => job_hash }
   #      # run_at is an epoch Integer.
   #    end
   #
-  class Workers
+  class WorkSet
     include Enumerable
-    include RedisScanner
 
-    def each
+    def each(&block)
+      results = []
       Sidekiq.redis do |conn|
-        procs = sscan(conn, "processes")
+        procs = conn.sscan_each("processes").to_a
         procs.sort.each do |key|
           valid, workers = conn.pipelined {
-            conn.exists(key)
+            conn.exists?(key)
             conn.hgetall("#{key}:workers")
           }
           next unless valid
           workers.each_pair do |tid, json|
-            yield key, tid, Sidekiq.load_json(json)
+            hsh = Sidekiq.load_json(json)
+            p = hsh["payload"]
+            # avoid breaking API, this is a side effect of the JSON optimization in #4316
+            hsh["payload"] = Sidekiq.load_json(p) if p.is_a?(String)
+            results << [key, tid, hsh]
           end
         end
       end
+
+      results.sort_by { |(_, _, hsh)| hsh["run_at"] }.each(&block)
     end
 
     # Note that #size is only as accurate as Sidekiq's heartbeat,
@@ -911,7 +984,7 @@ module Sidekiq
     # which can easily get out of sync with crashy processes.
     def size
       Sidekiq.redis do |conn|
-        procs = sscan(conn, "processes")
+        procs = conn.sscan_each("processes").to_a
         if procs.empty?
           0
         else
@@ -919,9 +992,13 @@ module Sidekiq
             procs.each do |key|
               conn.hget(key, "busy")
             end
-          }.map(&:to_i).inject(:+)
+          }.sum(&:to_i)
         end
       end
     end
   end
+  # Since "worker" is a nebulous term, we've deprecated the use of this class name.
+  # Is "worker" a process, a type of job, a thread? Undefined!
+  # WorkSet better describes the data.
+  Workers = WorkSet
 end