1
0
Fork 0
mirror of https://github.com/mperham/sidekiq.git synced 2022-11-09 13:52:34 -05:00
mperham--sidekiq/lib/sidekiq/api.rb
Mike Perham f220897db9
Per-job execution metrics (#5384)
* New execution stats data format, #5283

We store time and counts per-queue, per-class and totals, daily.

* break metrics into separate jobs and queues hashes

* tweak

* Move metrics tracking to middleware, start work on Query API

* Add support for labeled points in time

* Add fetch method for deploy marks

* rejigger metrics file layout

* fix tests

* Remove per-queue metrics, adds a lot of complexity with little value IMO

* store per-minute histograms

* Keep basic stats hardwired as is

The idea being that metrics are optional middleware as they have significantly more overhead in CPU time and Redis space.

* Implement top N metrics dashboard

* Add topN and job-specific metric pages

* Supply histogram data to job metrics page

* cant use local time as CI is in a different tz

* Add basic metrics graph, refactor dashboard JS to make Rickshaw reuseable

* prepare for public beta
2022-07-28 12:21:42 -07:00

1136 lines
30 KiB
Ruby

# frozen_string_literal: true
require "sidekiq"
require "zlib"
require "set"
require "base64"
require "sidekiq/metrics/deploy"
require "sidekiq/metrics/query"
module Sidekiq
# Retrieve runtime statistics from Redis regarding
# this Sidekiq cluster.
#
# stat = Sidekiq::Stats.new
# stat.processed
class Stats
def initialize
fetch_stats_fast!
end
def processed
stat :processed
end
def failed
stat :failed
end
def scheduled_size
stat :scheduled_size
end
def retry_size
stat :retry_size
end
def dead_size
stat :dead_size
end
def enqueued
stat :enqueued
end
def processes_size
stat :processes_size
end
def workers_size
stat :workers_size
end
def default_queue_latency
stat :default_queue_latency
end
def queues
Sidekiq::Stats::Queues.new.lengths
end
# O(1) redis calls
# @api private
def fetch_stats_fast!
pipe1_res = Sidekiq.redis { |conn|
conn.pipelined do |pipeline|
pipeline.get("stat:processed")
pipeline.get("stat:failed")
pipeline.zcard("schedule")
pipeline.zcard("retry")
pipeline.zcard("dead")
pipeline.scard("processes")
pipeline.lrange("queue:default", -1, -1)
end
}
default_queue_latency = if (entry = pipe1_res[6].first)
job = begin
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,
scheduled_size: pipe1_res[2],
retry_size: pipe1_res[3],
dead_size: pipe1_res[4],
processes_size: pipe1_res[5],
default_queue_latency: default_queue_latency
}
end
# O(number of processes + number of queues) redis calls
# @api private
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 |pipeline|
processes.each { |key| pipeline.hget(key, "busy") }
queues.each { |queue| pipeline.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
# @api private
def fetch_stats!
fetch_stats_fast!
fetch_stats_slow!
end
# @api private
def reset(*stats)
all = %w[failed processed]
stats = stats.empty? ? all : all & stats.flatten.compact.map(&:to_s)
mset_args = []
stats.each do |stat|
mset_args << "stat:#{stat}"
mset_args << 0
end
Sidekiq.redis do |conn|
conn.mset(*mset_args)
end
end
private
def stat(s)
fetch_stats_slow! if @stats[s].nil?
@stats[s] || raise(ArgumentError, "Unknown stat #{s}")
end
class Queues
def lengths
Sidekiq.redis do |conn|
queues = conn.sscan_each("queues").to_a
lengths = conn.pipelined { |pipeline|
queues.each do |queue|
pipeline.llen("queue:#{queue}")
end
}
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
def processed
@processed ||= date_stat_hash("processed")
end
def failed
@failed ||= date_stat_hash("failed")
end
private
def date_stat_hash(stat)
stat_hash = {}
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|
conn.mget(keys).each_with_index do |value, idx|
stat_hash[dates[idx]] = value ? value.to_i : 0
end
end
rescue RedisConnection.adapter::CommandError
# mget will trigger a CROSSSLOT error when run against a Cluster
# TODO Someone want to add Cluster support?
end
stat_hash
end
end
end
##
# Represents a queue within Sidekiq.
# Allows enumeration of all jobs within the queue
# and deletion of jobs. NB: this queue data is real-time
# and is changing within Redis moment by moment.
#
# queue = Sidekiq::Queue.new("mailer")
# queue.each do |job|
# job.klass # => 'MyWorker'
# job.args # => [1, 2, 3]
# job.delete if job.jid == 'abcdef1234567890'
# end
class Queue
include Enumerable
##
# Fetch all known queues within Redis.
#
# @return [Array<Sidekiq::Queue>]
def self.all
Sidekiq.redis { |c| c.sscan_each("queues").to_a }.sort.map { |q| Sidekiq::Queue.new(q) }
end
attr_reader :name
# @param name [String] the name of the queue
def initialize(name = "default")
@name = name.to_s
@rname = "queue:#{name}"
end
# The current size of the queue within Redis.
# This value is real-time and can change between calls.
#
# @return [Integer] the size
def size
Sidekiq.redis { |con| con.llen(@rname) }
end
# @return [Boolean] if the queue is currently paused
def paused?
false
end
##
# Calculates this queue's latency, the difference in seconds since the oldest
# job in the queue was enqueued.
#
# @return [Float] in seconds
def latency
entry = Sidekiq.redis { |conn|
conn.lrange(@rname, -1, -1)
}.first
return 0 unless entry
job = Sidekiq.load_json(entry)
now = Time.now.to_f
thence = job["enqueued_at"] || now
now - thence
end
def each
initial_size = size
deleted_size = 0
page = 0
page_size = 50
loop do
range_start = page * page_size - deleted_size
range_end = range_start + page_size - 1
entries = Sidekiq.redis { |conn|
conn.lrange @rname, range_start, range_end
}
break if entries.empty?
page += 1
entries.each do |entry|
yield JobRecord.new(entry, @name)
end
deleted_size = initial_size - size
end
end
##
# Find the job with the given JID within this queue.
#
# This is a *slow, inefficient* operation. Do not use under
# normal conditions.
#
# @param jid [String] the job_id to look for
# @return [Sidekiq::JobRecord]
# @return [nil] if not found
def find_job(jid)
detect { |j| j.jid == jid }
end
# delete all jobs within this queue
# @return [Boolean] true
def clear
Sidekiq.redis do |conn|
conn.multi do |transaction|
transaction.unlink(@rname)
transaction.srem("queues", name)
end
end
true
end
alias_method :💣, :clear
# :nodoc:
# @api private
def as_json(options = nil)
{name: name} # 5336
end
end
##
# Represents a pending job within a Sidekiq queue.
#
# The job should be considered immutable but may be
# removed from the queue via JobRecord#delete.
class JobRecord
# the parsed Hash of job data
# @!attribute [r] Item
attr_reader :item
# the underlying String in Redis
# @!attribute [r] Value
attr_reader :value
# the queue associated with this job
# @!attribute [r] Queue
attr_reader :queue
# :nodoc:
# @api private
def initialize(item, queue_name = nil)
@args = nil
@value = item
@item = item.is_a?(Hash) ? item : parse(item)
@queue = queue_name || @item["queue"]
end
# :nodoc:
# @api private
def parse(item)
Sidekiq.load_json(item)
rescue JSON::ParserError
# If the job payload in Redis is invalid JSON, we'll load
# the item as an empty hash and store the invalid JSON as
# the job 'args' for display in the Web UI.
@invalid = true
@args = [item]
{}
end
# This is the job class which Sidekiq will execute. If using ActiveJob,
# this class will be the ActiveJob adapter class rather than a specific job.
def klass
self["class"]
end
def display_class
# Unwrap known wrappers so they show up in a human-friendly manner in the Web UI
@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
def display_args
# Unwrap known wrappers so they show up in a human-friendly manner in the Web UI
@display_args ||= case klass
when /\ASidekiq::Extensions::Delayed/
safe_load(args[0], args) do |_, _, arg, kwarg|
if !kwarg || kwarg.empty?
arg
else
[arg, kwarg]
end
end
when "ActiveJob::QueueAdapters::SidekiqAdapter::JobWrapper"
job_args = self["wrapped"] ? args[0]["arguments"] : []
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
else
if self["encrypt"]
# no point in showing 150+ bytes of random garbage
args[-1] = "[encrypted data]"
end
args
end
end
def args
@args || @item["args"]
end
def jid
self["jid"]
end
def enqueued_at
self["enqueued_at"] ? Time.at(self["enqueued_at"]).utc : nil
end
def created_at
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
def latency
now = Time.now.to_f
now - (@item["enqueued_at"] || @item["created_at"] || now)
end
# Remove this job from the queue
def delete
count = Sidekiq.redis { |conn|
conn.lrem("queue:#{@queue}", 1, @value)
}
count != 0
end
# Access arbitrary attributes within the job hash
def [](name)
# nil will happen if the JSON fails to parse.
# We don't guarantee Sidekiq will work with bad job JSON but we should
# make a best effort to minimize the damage.
@item ? @item[name] : nil
end
private
def safe_load(content, default)
yield(*YAML.load(content))
rescue => ex
# #1761 in dev mode, it's possible to have jobs enqueued which haven't been loaded into
# memory yet so the YAML can't be loaded.
# TODO is this still necessary? Zeitwerk reloader should handle?
Sidekiq.logger.warn "Unable to load YAML: #{ex.message}" unless Sidekiq.config[: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
# Represents a job within a Redis sorted set where the score
# represents a timestamp associated with the job. This timestamp
# could be the scheduled time for it to run (e.g. scheduled set),
# or the expiration date after which the entry should be deleted (e.g. dead set).
class SortedEntry < JobRecord
attr_reader :score
attr_reader :parent
# :nodoc:
# @api private
def initialize(parent, score, item)
super(item)
@score = Float(score)
@parent = parent
end
# The timestamp associated with this entry
def at
Time.at(score).utc
end
# remove this entry from the sorted set
def delete
if @value
@parent.delete_by_value(@parent.name, @value)
else
@parent.delete_by_jid(score, jid)
end
end
# Change the scheduled time for this job.
#
# @param at [Time] the new timestamp for this job
def reschedule(at)
Sidekiq.redis do |conn|
conn.zincrby(@parent.name, at.to_f - @score, Sidekiq.dump_json(@item))
end
end
# Enqueue this job from the scheduled or dead set so it will
# be executed at some point in the near future.
def add_to_queue
remove_job do |message|
msg = Sidekiq.load_json(message)
Sidekiq::Client.push(msg)
end
end
# enqueue this job from the retry set so it will be executed
# at some point in the near future.
def retry
remove_job do |message|
msg = Sidekiq.load_json(message)
msg["retry_count"] -= 1 if msg["retry_count"]
Sidekiq::Client.push(msg)
end
end
# Move this job from its current set into the Dead set.
def kill
remove_job do |message|
DeadSet.new.kill(message)
end
end
def error?
!!item["error_class"]
end
private
def remove_job
Sidekiq.redis do |conn|
results = conn.multi { |transaction|
transaction.zrangebyscore(parent.name, score, score)
transaction.zremrangebyscore(parent.name, score, score)
}.first
if results.size == 1
yield results.first
else
# multiple jobs with the same score
# find the one with the right JID and push it
matched, nonmatched = results.partition { |message|
if message.index(jid)
msg = Sidekiq.load_json(message)
msg["jid"] == jid
else
false
end
}
msg = matched.first
yield msg if msg
# push the rest back onto the sorted set
conn.multi do |transaction|
nonmatched.each do |message|
transaction.zadd(parent.name, score.to_f.to_s, message)
end
end
end
end
end
end
# Base class for all sorted sets within Sidekiq.
class SortedSet
include Enumerable
# Redis key of the set
# @!attribute [r] Name
attr_reader :name
# :nodoc:
# @api private
def initialize(name)
@name = name
@_size = size
end
# real-time size of the set, will change
def size
Sidekiq.redis { |c| c.zcard(name) }
end
# Scan through each element of the sorted set, yielding each to the supplied block.
# Please see Redis's <a href="https://redis.io/commands/scan/">SCAN documentation</a> for implementation details.
#
# @param match [String] a snippet or regexp to filter matches.
# @param count [Integer] number of elements to retrieve at a time, default 100
# @yieldparam [Sidekiq::SortedEntry] each entry
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
# @return [Boolean] always true
def clear
Sidekiq.redis do |conn|
conn.unlink(name)
end
true
end
alias_method :💣, :clear
# :nodoc:
# @api private
def as_json(options = nil)
{name: name} # 5336
end
end
# Base class for all sorted sets which contain jobs, e.g. scheduled, retry and dead.
# Sidekiq Pro and Enterprise add additional sorted sets which do not contain job data,
# e.g. Batches.
class JobSet < SortedSet
# Add a job with the associated timestamp to this set.
# @param timestamp [Time] the score for the job
# @param job [Hash] the job data
def schedule(timestamp, job)
Sidekiq.redis do |conn|
conn.zadd(name, timestamp.to_f.to_s, Sidekiq.dump_json(job))
end
end
def each
initial_size = @_size
offset_size = 0
page = -1
page_size = 50
loop do
range_start = page * page_size + offset_size
range_end = range_start + page_size - 1
elements = Sidekiq.redis { |conn|
conn.zrange name, range_start, range_end, withscores: true
}
break if elements.empty?
page -= 1
elements.reverse_each do |element, score|
yield SortedEntry.new(self, score, element)
end
offset_size = initial_size - @_size
end
end
##
# Fetch jobs that match a given time or Range. Job ID is an
# optional second argument.
#
# @param score [Time,Range] a specific timestamp or range
# @param jid [String, optional] find a specific JID within the score
# @return [Array<SortedEntry>] any results found, can be empty
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, begin_score, end_score, withscores: true)
}
elements.each_with_object([]) do |element, result|
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 O(n) operation*. Do not use for app logic.
#
# @param jid [String] the job identifier
# @return [SortedEntry] the record or nil
def find_job(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
# :nodoc:
# @api private
def delete_by_value(name, value)
Sidekiq.redis do |conn|
ret = conn.zrem(name, value)
@_size -= 1 if ret
ret
end
end
# :nodoc:
# @api private
def delete_by_jid(score, jid)
Sidekiq.redis do |conn|
elements = conn.zrangebyscore(name, score, score)
elements.each do |element|
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
end
alias_method :delete, :delete_by_jid
end
##
# The set of scheduled jobs within Sidekiq.
# Based on this, you can search/filter for jobs. Here's an
# example where I'm selecting jobs based on some complex logic
# and deleting them from the scheduled set.
#
# r = Sidekiq::ScheduledSet.new
# r.select do |scheduled|
# scheduled.klass == 'Sidekiq::Extensions::DelayedClass' &&
# scheduled.args[0] == 'User' &&
# scheduled.args[1] == 'setup_new_subscriber'
# end.map(&:delete)
class ScheduledSet < JobSet
def initialize
super "schedule"
end
end
##
# The set of retries within Sidekiq.
# Based on this, you can search/filter for jobs. Here's an
# example where I'm selecting all jobs of a certain type
# and deleting them from the retry queue.
#
# r = Sidekiq::RetrySet.new
# r.select do |retri|
# retri.klass == 'Sidekiq::Extensions::DelayedClass' &&
# retri.args[0] == 'User' &&
# retri.args[1] == 'setup_new_subscriber'
# end.map(&:delete)
class RetrySet < JobSet
def initialize
super "retry"
end
# Enqueues all jobs pending within the retry set.
def retry_all
each(&:retry) while size > 0
end
# Kills all jobs pending within the retry set.
def kill_all
each(&:kill) while size > 0
end
end
##
# The set of dead jobs within Sidekiq. Dead jobs have failed all of
# their retries and are helding in this set pending some sort of manual
# fix. They will be removed after 6 months (dead_timeout) if not.
#
class DeadSet < JobSet
def initialize
super "dead"
end
# Add the given job to the Dead set.
# @param message [String] the job data as JSON
def kill(message, opts = {})
now = Time.now.to_f
Sidekiq.redis do |conn|
conn.multi do |transaction|
transaction.zadd(name, now.to_s, message)
transaction.zremrangebyscore(name, "-inf", now - self.class.timeout)
transaction.zremrangebyrank(name, 0, - self.class.max_jobs)
end
end
if opts[:notify_failure] != false
job = Sidekiq.load_json(message)
r = RuntimeError.new("Job killed by API")
r.set_backtrace(caller)
Sidekiq.death_handlers.each do |handle|
handle.call(job, r)
end
end
true
end
# Enqueue all dead jobs
def retry_all
each(&:retry) while size > 0
end
# The maximum size of the Dead set. Older entries will be trimmed
# to stay within this limit. Default value is 10,000.
def self.max_jobs
Sidekiq[:dead_max_jobs]
end
# The time limit for entries within the Dead set. Older entries will be thrown away.
# Default value is six months.
def self.timeout
Sidekiq[:dead_timeout_in_seconds]
end
end
##
# Enumerates the set of Sidekiq processes which are actively working
# right now. Each process sends a heartbeat to Redis every 5 seconds
# so this set should be relatively accurate, barring network partitions.
#
# @yieldparam [Sidekiq::Process]
#
class ProcessSet
include Enumerable
# :nodoc:
# @api private
def initialize(clean_plz = true)
cleanup if clean_plz
end
# Cleans up dead processes recorded in Redis.
# Returns the number of processes cleaned.
# :nodoc:
# @api private
def cleanup
return 0 unless Sidekiq.redis { |conn| conn.set("process_cleanup", "1", nx: true, ex: 60) }
count = 0
Sidekiq.redis do |conn|
procs = conn.sscan_each("processes").to_a.sort
heartbeats = conn.pipelined { |pipeline|
procs.each do |key|
pipeline.hget(key, "info")
end
}
# 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 = procs.select.with_index { |proc, i|
heartbeats[i].nil?
}
count = conn.srem("processes", to_prune) unless to_prune.empty?
end
count
end
def each
result = Sidekiq.redis { |conn|
procs = conn.sscan_each("processes").to_a.sort
# 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
conn.pipelined do |pipeline|
procs.each do |key|
pipeline.hmget(key, "info", "busy", "beat", "quiet", "rss", "rtt_us")
end
end
}
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,
"rss" => rss.to_i,
"rtt_us" => rtt.to_i))
end
end
# This method is not guaranteed accurate since it does not prune the set
# based on current heartbeat. #each does that and ensures the set only
# contains Sidekiq processes which have sent a heartbeat within the last
# 60 seconds.
# @return [Integer] current number of registered Sidekiq processes
def size
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.
# @return [Integer] the sum of process concurrency
def total_concurrency
sum { |x| x["concurrency"].to_i }
end
# @return [Integer] total amount of RSS memory consumed by Sidekiq processes
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.
# @return [String] Identity of cluster leader
# @return [String] empty string if no leader
def leader
@leader ||= begin
x = Sidekiq.redis { |c| c.get("dear-leader") }
# need a non-falsy value so we can memoize
x ||= ""
x
end
end
end
#
# Sidekiq::Process represents an active Sidekiq process talking with Redis.
# Each process has a set of attributes which look like this:
#
# {
# 'hostname' => 'app-1.example.com',
# 'started_at' => <process start time>,
# 'pid' => 12345,
# 'tag' => 'myapp'
# 'concurrency' => 25,
# 'queues' => ['default', 'low'],
# 'busy' => 10,
# 'beat' => <last heartbeat>,
# 'identity' => <unique string identifying the process>,
# }
class Process
# :nodoc:
# @api private
def initialize(hash)
@attribs = hash
end
def tag
self["tag"]
end
def labels
Array(self["labels"])
end
def [](key)
@attribs[key]
end
def identity
self["identity"]
end
def queues
self["queues"]
end
# Signal this process to stop processing new jobs.
# It will continue to execute jobs it has already fetched.
# This method is *asynchronous* and it can take 5-10
# seconds for the process to quiet.
def quiet!
signal("TSTP")
end
# Signal this process to shutdown.
# It will shutdown within its configured :timeout value, default 25 seconds.
# This method is *asynchronous* and it can take 5-10
# seconds for the process to start shutting down.
def stop!
signal("TERM")
end
# Signal this process to log backtraces for all threads.
# Useful if you have a frozen or deadlocked process which is
# still sending a heartbeat.
# This method is *asynchronous* and it can take 5-10 seconds.
def dump_threads
signal("TTIN")
end
# @return [Boolean] true if this process is quiet or shutting down
def stopping?
self["quiet"] == "true"
end
private
def signal(sig)
key = "#{identity}-signals"
Sidekiq.redis do |c|
c.multi do |transaction|
transaction.lpush(key, sig)
transaction.expire(key, 60)
end
end
end
end
##
# The WorkSet stores the work being done by this Sidekiq cluster.
# It tracks the process and thread working on each job.
#
# WARNING WARNING WARNING
#
# This is live data that can change every millisecond.
# If you call #size => 5 and then expect #each to be
# called 5 times, you're going to have a bad time.
#
# 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' => job_hash }
# # run_at is an epoch Integer.
# end
#
class WorkSet
include Enumerable
def each(&block)
results = []
Sidekiq.redis do |conn|
procs = conn.sscan_each("processes").to_a
procs.sort.each do |key|
valid, workers = conn.pipelined { |pipeline|
pipeline.exists?(key)
pipeline.hgetall("#{key}:work")
}
next unless valid
workers.each_pair do |tid, 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,
# which happens every 5 seconds. It is NOT real-time.
#
# Not very efficient if you have lots of Sidekiq
# processes but the alternative is a global counter
# which can easily get out of sync with crashy processes.
def size
Sidekiq.redis do |conn|
procs = conn.sscan_each("processes").to_a
if procs.empty?
0
else
conn.pipelined { |pipeline|
procs.each do |key|
pipeline.hget(key, "busy")
end
}.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