I cannot get apn_sender to use my RedisToGo url, it uses the default localhost regardless of me setting this in a resque.rb initializer:

if ENV["REDISTOGO_URL"] && uri = URI.parse(ENV["REDISTOGO_URL"])
  Resque.redis = Redis.new(host: uri.host, port: uri.port, password: uri.password)
end

I have this in my Procfile on Heroku:

apn: bundle exec rake apn:sender

*** Starting worker to send apple notifications in the background from Fredriks-iMac.local:53010:apple_push_notifications

How does this send notifications through the same persistent socket? If I queue notifications and run a worker in dev then step through it in the debugger, setup_connection is always called when the socket is written to. The @socket ivar is always nil, I assume because jobs are forked before they are run so how is the socket method in APN::Connection:Base supposed to evaluate?

      def socket
        setup_connection unless @socket
        return @socket
      end

PS. Great library!

Push notifications work fine, until there is some period of inactivity.

After that, the job raises the following errors:

An Errno::EPIPE occurred in background at 2013-10-13 15:39:30 UTC :

Broken pipe
/usr/lib/ruby/1.9.1/openssl/buffering.rb:317:in `syswrite'

It seems the pool timed out, but the job doesn't try to reconnect.

i try to send words with umlauts (ä,ö,ü) but i get an rails error while sending:

.../push_it/app/controllers/push_controller.rb:4: invalid multibyte char (US-ASCII)
.../push_it/app/controllers/push_controller.rb:4: invalid multibyte char (US-ASCII)
.../push_it/app/controllers/push_controller.rb:4: syntax error, unexpected $end, expecting ')'
...ch unser Weihachtsspecial veröffentlicht. Schaut es euch an...           

Hey,
I'm been trying to generate the apn_sender daemon script this way:

./script/generate apn_sender

but it returns "zsh: no such file or directory: ./script/generate"

Using the rails way to generate stuffs:

rails g apn_sender

returns "Could not find generator apn_sender."

Rails version: Rails 3.2.13
ruby version: ruby 1.9.3p194 (2012-04-20 revision 35410) [x86_64-darwin12.0.0]

Do you have an idea why it doesn't work?
Thanks in advance for your help

Let's look at the example from apn/tasks:

APN.password = ENV['CERT_PASS']
APN.full_certificate_path =  ENV['FULL_CERT_PATH']
APN.logger = Rails.logger

worker = ::Resque::Worker.new(APN::Jobs::QUEUE_NAME)

puts "*** Starting worker to send apple notifications in the background from #{worker}"
worker.work(ENV['INTERVAL'] || 5) # interval, will block

This starts just fine. However...

• require 'apn/jobs/resque_notification_job' is now run inside the backend method.
• The queue uses the job class APN::Jobs::ResqueNotificationJob.

So, since we've never called APN.backend in this worker (and we shouldn't have to do that!), the first queue item will crash when attempting to reference the un-loaded ResqueNotificationJob class.

I think the solution should load the job classes up front, rather than requiring a backend (which the workers won't even use) to be created.

How do you feel about trying to require both sidekiq and resque, like below?

begin
  require 'sidekiq'
  require 'apn/jobs/sidekiq_notification_job'
rescue LoadError
end

begin
  require 'resque'
  require 'apn/jobs/resque_notification_job'
rescue LoadError
end

This solves my problems, although it feels a little less than clean. If it's fine with you, I'll send a pull request.

Steps to reproduce:

1. Launch APN sender in a loop where the connection pool is persisted despite exceptions (i.e., Sidekiq).
2. Start sending notifications.
3. Simulate network problems.

Expected behaviour: Broken connections are repaired or discarded, failed messages are possibly retried.
Observed behaviour: Broken connections are supplied for the next messages.

Where is it recommended to put the configuration settings?

Also, is it possible to support pushing to multiple different applications (i.e. iphone and ipad versions)?

Hi,

I try to use the APN feedback to get the invalid tokens, so that we can remove them from the database.

But this is what I get on the console in production:

irb(main):011:0>  feedback = APN::Feedback.new()
=> #
irb(main):004:0>  tokens = feedback.tokens
=> []

The server sends push with notify_sync and notify_async. So that is working.
There are invalid tokens on my database, so the array should not be empty.

Thanks.

Hi, guys, so I have question. What is the rigth way to send push notifications to array of tokens ?
Example, input array ["token1", "token2", "token3"]
What I do

["token1", "token2", "token3"].each do |token|
   APN.notify_async(token, options blah blah)
end

I think this is not a right way to send multiple notifications. Because each time created new persistent connection to apn socket. Or connection always opened ? Please correct me.

We need to add the new features that are on version 2.0:

• sidekiq
• multithread
• connection pooling

I'm using rails and heroku. It works perfectly as apn sender. But I got an error in trying to use "heroku run console".

/app/vendor/ruby-2.0.0/lib/ruby/2.0.0/open-uri.rb:36:in `initialize': No such file or directory - /app/log/apn_sender.log (Errno::ENOENT)
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/open-uri.rb:36:in `open'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/open-uri.rb:36:in `open'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/logger.rb:601:in `create_logfile'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/logger.rb:596:in `open_logfile'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/logger.rb:551:in `initialize'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/logger.rb:318:in `new'
        from /app/vendor/ruby-2.0.0/lib/ruby/2.0.0/logger.rb:318:in `initialize'
        from /app/vendor/bundle/ruby/2.0.0/gems/activesupport-3.2.16/lib/active_support/core_ext/logger.rb:72:in `initialize'
        from /app/vendor/bundle/ruby/2.0.0/gems/apn_sender-2.0.1/lib/apn/railtie.rb:11:in `new'
        from /app/vendor/bundle/ruby/2.0.0/gems/apn_sender-2.0.1/lib/apn/railtie.rb:11:in `block in '
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/initializable.rb:30:in `instance_exec'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/initializable.rb:30:in `run'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/initializable.rb:55:in `block in run_initializers'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/initializable.rb:54:in `each'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/initializable.rb:54:in `run_initializers'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/application.rb:136:in `initialize!'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/railtie/configurable.rb:30:in `method_missing'
        from /app/config/environment.rb:10:in `'
        from /app/vendor/bundle/ruby/2.0.0/gems/activesupport-3.2.16/lib/active_support/dependencies.rb:251:in `require'
        from /app/vendor/bundle/ruby/2.0.0/gems/activesupport-3.2.16/lib/active_support/dependencies.rb:251:in `block in require'
        from /app/vendor/bundle/ruby/2.0.0/gems/activesupport-3.2.16/lib/active_support/dependencies.rb:236:in `load_dependency'
        from /app/vendor/bundle/ruby/2.0.0/gems/activesupport-3.2.16/lib/active_support/dependencies.rb:251:in `require'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/application.rb:103:in `require_environment!'
        from /app/vendor/bundle/ruby/2.0.0/gems/railties-3.2.16/lib/rails/commands.rb:40:in `'
        from script/rails:6:in `require'
        from script/rails:6:in `'

here is my .gitignore and .slugignore

> cat .gitignore
.bundle
db/*.sqlite3
log/*.log
tmp/
.sass-cache/
coverage/
.envrc
tags
.project
*~

> cat .slugignore
coverage/
doc/
*.log
spec/
test/
README.rst

Is there no way to use "heroku run console" with apn_sender?

First of all, sorry if this is a noob question, but I've spent a lot of time with this without success. I set up this with resque, and all goes right if I follow your default configuration. But when I'm trying to change the APN.password it throws an error. The same happens if I try to change the cert_name.

I put on config/environments/production.rb but I think it's not the right place since who send those notifications are the resque workers.

Thank you in advance and thank you for this great gem!

Resque.enqueue(APN::Jobs::ResqueNotificationJob, token, opts) is being called and a job is being scheduled whenever i execute a code APN.notify_async(token, opts) from my rails application in development environment. This job is successfully getting processed with no failure as shown at url /resque/stats/resque
But the notification is not being pushed and no output is logged at 'logs/apn_sender.log'.

Then i tried to send push notification directly from rails console by calling: APN::Jobs::ResqueNotificationJob.perform(token, opts)
This do send push notification.

Hi there,

Apple introduced a new byte format of messages that supports up to 2048 byte limit for messages etc. and a lot more. Be careful, the byte format of these new messages is different. So just sending more bytes via old commands does not work, regardless whether the apn sender limitation within this code has been increased.

Compare this here Figure 5-1 from here (the new one - command level 2):
https://developer.apple.com/library/mac/documentation/NetworkingInternet/Conceptual/RemoteNotificationsPG/Chapters/CommunicatingWIthAPS.html#//apple_ref/doc/uid/TP40008194-CH101-SW4

to here Figure A -1 (very old - command level 0):
https://developer.apple.com/library/mac/documentation/NetworkingInternet/Conceptual/RemoteNotificationsPG/Chapters/LegacyFormat.html#//apple_ref/doc/uid/TP40008194-CH105-SW2

and here Figure A - 2 (newer - command level 1)
https://developer.apple.com/library/mac/documentation/NetworkingInternet/Conceptual/RemoteNotificationsPG/Chapters/LegacyFormat.html#//apple_ref/doc/uid/TP40008194-CH105-SW5

The last two are (now) flagged as legacy and anything send via this format gets treated accordingly (incl limitations)
Only the newest one supports the larger messages and these other stuff.

Let's see whether we find the time to open a pull request :)
Cheers
SH

Hello guys, i'm having a bit of trouble in sending those push notifications. When "notify_async(ids,opts)" is called i get the following error:

NoMethodError (undefined method `notify' for :sidekiq:Symbol):

Here's the screenshot of the code line

def self.send_apple_push_notification(projeto, ids, message)
 projeto.configure_apn(projeto)
 options = { alert: message, badge: "+1", sound: true, other_information: ''}
 ids.each do |id|
   APN.notify_async(id, options)
 end
end

I've chosen sidekiq as my backend just like the README suggests

APN.backend = :sidekiq # use sidekiq backend

Here's the rest of the configuration. This was done with the intent to make the project work with multiple certificates.

def configure_apn(projeto)
  path = nil
  if projeto.desenvolvimento
    path = projeto.apn_development.path
  else
    path = projeto.apn_production.path
  end

  unless path.nil?
    last_slash = path.rindex("/")
    folder = path[0..last_slash]
    file = path[last_slash+1..path.length]

    APN.root = folder # root to certificates folder
    APN.certificate_name = file # certificate filename
    if projeto.desenvolvimento
      APN.host = projeto.apn_host
      APN.password = projeto.apn_password_development
    else
      APN.password = projeto.apn_password_production
    end
    APN.pool_size = 1 # number of connections on the pool
    APN.pool_timeout = 5 # timeout in seconds for connection pool
    APN.logger = Logger.new(File.join(Rails.root, 'log', 'apn_sender.log'))
    APN.truncate_alert = true 
    APN.backend = :sidekiq # use sidekiq backend
  end
end

Heroku (and localhost if tmp/pids is not present) bombs:

 /app/vendor/bundle/ruby/1.9.1/gems/daemons-1.1.9/lib/daemons/pidfile.rb:94:in `initialize': No such file or directory - /app/tmp/pids/apn_sender.pid (Errno::ENOENT)

Try to used the apn_sender from rubygems. Got following error. install apn_sender from github works for me. time for a release?

*** DEPRECATION WARNING: Resque::Worker#verbose and #very_verbose are deprecated. Please set Resque.logger.level instead
Called from: /home/ /.rvm/gems/ruby-1.9.3-p429/gems/resque-1.24.1/lib/resque/worker.rb:649:in verbose=' /home/ /.rvm/gems/ruby-1.9.3-p429/gems/apn_sender-1.0.6/lib/apn/tasks.rb:12:inblock (2 levels) in <top (required)>'
/home/ /.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:246:in call' /home/ /.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:246:inblock in execute'
/home/ /.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:241:in each' /home/ /.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:241:inexecute'
*** Starting worker to send apple notifications in the background from ip-10-48-241-94:3349:apple_push_notifications
WARNING: This way of doing signal handling is now deprecated. Please see http://hone.heroku.com/resque/2012/08/21/resque-signals.html for more info.
I, [2013-06-24T14:11:59.538314 #3349] INFO -- : 2013-06-24 14:11:59 +0000: Running before_first_fork hook
I, [2013-06-24T14:11:59.538823 #3349] INFO -- : 2013-06-24 14:11:59 +0000: Failed to start worker : #<NoMethodError: undefined method call' for []:Array> rake aborted! wrong number of arguments (1 for 0) /home//.rvm/gems/ruby-1.9.3-p429/gems/apn_sender-1.0.6/lib/resque/hooks/before_unregister_worker.rb:10:in unregister_worker'
/home/ /.rvm/gems/ruby-1.9.3-p429/gems/resque-1.24.1/lib/resque/worker.rb:176:inrescue in work' /home//.rvm/gems/ruby-1.9.3-p429/gems/resque-1.24.1/lib/resque/worker.rb:130:in work'
/home//.rvm/gems/ruby-1.9.3-p429/gems/apn_sender-1.0.6/lib/apn/tasks.rb:17:inblock (2 levels) in <top (required)>' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:246:in call'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:246:inblock in execute' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:241:in each'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:241:inexecute' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:184:in block in invoke_with_call_chain'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:177:ininvoke_with_call_chain' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/task.rb:170:in invoke'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:143:ininvoke_task' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:101:in block (2 levels) in top_level'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:101:ineach' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:101:in block in top_level'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:110:inrun_with_threads' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:95:in top_level'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:73:inblock in run' /home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:160:in standard_exception_handling'
/home//.rvm/gems/ruby-1.9.3-p429@global/gems/rake-10.0.4/lib/rake/application.rb:70:inrun' /home//.rvm/gems/ruby-1.9.3-p429/bin/ruby_noexec_wrapper:14:in eval'
/home//.rvm/gems/ruby-1.9.3-p429/bin/ruby_noexec_wrapper:14:in`

Any plans for a version bump? Need help with anything to get there?

Just trying to use apn_sender with sidekiq in development mode:

APN.host = 'gateway.sandbox.push.apple.com'
APN.password = 'justapw'
APN.pool_size = 1 # number of connections on the pool
APN.pool_timeout = 5 # timeout in seconds for connection pool
device_token = "123"
APN.backend = :sidekiq
APN.notify_async(device_token, :alert => 'PushPlugin works!!', :badge => 1, :sound => 'beep.wav')

And sidekiq askes me for a password?

2014-01-31T17:48:52Z 12372 TID-ko1ws APN::Jobs::SidekiqNotificationJob JID-11111 INFO: start
Enter PEM pass phrase:

Could you please integrate changes from several forks that have a working Rails.env and Rails.logger? It is very hard to debug anything on rails 3 without those actually working

Hi Kali,

I'm testin apn_sender with resque, in my sandbox work fine, the notifications are sent and received in my iphone, but in Heroku the notifications not arrive to my iphone

The resque jobs are working, in the log I see when the #perform method when proper values are executed, no errors, but the notificactions don't arrive to my iphone

Any clue what I'm doing wrong?

In my environment the throughput can be as low as 8 notification send/s. Just wondering is this typical performance?

→ rails g apn_sender
Could not find generator apn_sender.

Please help!

The size of the payload is computed with packaged_notification.size.to_i which includes [0, 0, 32, pt, 0, pm.size, pm]. Apple documentation (see https://developer.apple.com/library/mac/documentation/NetworkingInternet/Conceptual/RemoteNotificationsPG/Chapters/CommunicatingWIthAPS.html#//apple_ref/doc/uid/TP40008194-CH101-SW4) states that the payload size is the size of the json, not the size of the TCP packet.

I see that the majority of the messages are failing with an error:

OpenSSL::SSL::SSLError: SSL_write: bad write retry

I see the logs and I see that the apn_sender log stops like 3 days ago which afaik means that the worker died? Have you seen that behavior again? Can you think of any workaround?

I am in version 2.0.0

see https://github.com/rails/activejob for more info

Hi,
Does you gem support mdm for ios? If yes, can you please explain how to use it? If not, can you please add it?

Thanks.

is this a typo

Launching rake apn:sender with resque 1.24 gives this error :

/Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/gems/apn_sender-1.0.5/lib/resque/hooks/before_unregister_worker.rb:10:in unregister_worker' /Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/gems/resque-1.24.1/lib/resque/worker.rb:176:inrescue in work'
/Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/gems/resque-1.24.1/lib/resque/worker.rb:130:in work' /Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/gems/apn_sender-1.0.5/lib/apn/tasks.rb:24:inblock (2 levels) in <top (required)>'
/Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/bin/ruby_noexec_wrapper:14:in eval' /Users/Wizzard/.rvm/gems/ruby-1.9.3-p429/bin/ruby_noexec_wrapper:14:in

I have an odd issue. I have setup everything as per docs. My Gemfile contains:

gem 'apn_sender', github: 'arthurnn/apn_sender', branch: 'master'
gem 'sidekiq'

And regardless, Sidekiq is not being used. The backend is set to APN::Backend::Simple

I am testing via rake tests

my Gemfile needed:

require: ['apn', 'apn/jobs/sidekiq_notification_job']

Using sidekiq (2.15.2)

this is not the case in the 2.0.0 rubygems version.

I'm curious about the reason behind this commit: bd6b790

Why isn't activesupport >= 4.1.0 permitted?

It would be really useful if apn_sender could maintain persistent connections (as needed) to both the sandbox and production servers. This is useful for specific testing scenarios.

Maybe notify_async could have a :sandbox => true/false argument?

https://github.com/arthurnn/apn_sender/blob/master/lib/apn/connection.rb#L5

As you probably know, Sidekiq runs in a multithreaded environment. I believe calling APN.notify_sync from multiple threads will still cause a race condition for the conditional assignment and creation of the client.

    def connection_pool
      @pool ||= ConnectionPool.new(size: (pool_size || 1), timeout: (pool_timeout || 5)) do
        APN::Client.new(host: host,
                        port: port,
                        certificate: certificate,
                        password: password)
      end
    end

    def with_connection(&block)
      connection_pool.with(&block)
    end

APN::Client itself is already lazily opening a socket connect, so in my opinion it does not hurt to initialize the ConnectionPool in an earlier stage.

Another interesting idea is to see if the APN::Client could be a Celluloid::Actor of its own.

I generated the key with pass phrase. now I got the pem file. and the notification can be sent.

now my issue is, it asked me for

Enter PEM pass phrase:  

when I send notification or get feedback. I have the pass phrase, how to set the pass phrase as default? so no need to type it each time? thanks for your help.

For iOS7 remote-notification background fetch (application:didReceiveRemoteNotification:fetchCompletionHandler)

{ content-available: 1 } must be in aps hash.

https://developer.apple.com/library/ios/documentation/NetworkingInternet/Conceptual/RemoteNotificationsPG/Chapters/ApplePushService.html

see Table 3-1 Keys and values of the aps dictionary

when i execute
feedback= APN::Feedback.new(:environment => :development)

i get
=> #<APN::Feedback: Connection not currently established to feedback.sandbox.push.apple.com on 2196>

i alse have rake taks running
rake apn:sender RAILS_ENV=development

And neither can you add them to a custom queue using this:
module APN::Jobs
QUEUE_NAME = :your_custom_queue_name
end

Say I have 1 server running several rails apps.

Each rails app has it's own Apple Push Notification messages, each bound for a different application, thus different PEM certificates.

It seems that apn_sender doesn't start multiple times, nor can I specify during APN.notify which PEM certificate to use.

I'm more than willing to help mod, but I could use some guidance as to how you'd like me to do it (before I just slam a bunch of keys into your redis namespace)

error

We have noticed that when certain special characters are used, specifically apostrophe (’) and ellipsis (…), the notification does not get delivered to the device. The notifications show in the redis log below:

https://gist.github.com/875746

You can see that the elipses goes in as \u2026 and comes back out as \xe2\x80\xa6. I'm not sure if this is a Redis issue or something that can be done from the apn_sender side.

Have you heard or experienced anything like this or have a hint as to where I could look in the source to test with some of these characters?

On some notifications, I'm getting this error, on 2.0.1

A NameError occurred in background at 2014-01-10 19:06:41 UTC : 

uninitialized constant APN::Jobs::SidekiqNotificationJob 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/activesupport-3.2.14/lib/active_support/inflector/methods.rb:230:in `block in constantize' 

------------------------------- 
Backtrace: 
------------------------------- 

/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/activesupport-3.2.14/lib/active_support/inflector/methods.rb:230:in `block in constantize' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/activesupport-3.2.14/lib/active_support/inflector/methods.rb:229:in `each' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/activesupport-3.2.14/lib/active_support/inflector/methods.rb:229:in `constantize' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/activesupport-3.2.14/lib/active_support/core_ext/string/inflections.rb:54:in `constantize' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/sidekiq-2.17.1/lib/sidekiq/processor.rb:43:in `block in process' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/calls.rb:25:in `call' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/calls.rb:25:in `public_send' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/calls.rb:25:in `dispatch' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/calls.rb:67:in `dispatch' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/future.rb:14:in `block in new' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/thread_handle.rb:13:in `block in initialize' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/internal_pool.rb:100:in `call' 
/home/imaginary/lenshare2.imaginarycloud.com/shared/bundle/ruby/1.9.1/gems/celluloid-0.15.2/lib/celluloid/internal_pool.rb:100:in `block in create' 

------------------------------- 
Data: 
------------------------------- 

* data: {:message=> 
{"retry"=>true, 
"queue"=>"default", 
"class"=>"APN::Jobs::SidekiqNotificationJob", 
"args"=> 
["343d6a73d8e39bc988a995215a72d0f8c3d78754a4c1c3a8c477ded774336ccf", 
{"alert"=>"Miguel commented an instant 18:17: Fab", 
"sound"=>true, 
"badge"=>5}], 
"jid"=>"71ded2ece0f7f9917f4dba2b", 
"enqueued_at"=>1389380801.0102606}} 

I'm getting strange results from the feedback service. Most of the tokens are 140 chars long instead of 64. The timestamp is totally off sometime, like some date in 1989. Sometimes the Token is just "3b" and the timestamp: 2066-01-15T05:27:03-08:00

I'm just doing this:

feedback_data = APN::Feedback.new(:environment => :production).data
feedback_data.each do |item|

accessing item.token and item.timestamp here

end

Any idea what's the problem here? Thanks in advance!

Hi,

Messages are sent on alternations. I'm using sidekiq and this is what happens with notify async:

Message 1 - worker connects to APN - message arrives to the iphone
Message 2 - worker is connect, apn log says it was sent, never arrives
Message 3 - worker reconnects to APN - message arrives to the iphone
Message 4 - worker is connect, apn log says it was sent, never arrives
Message 5 - worker reconnects to APN - message arrives to the iphone

and so on.

When I try to start the daemon on my server it will tell me:
apn_sender: process with pid 1451 started.

But it is actually not running; "apn_sender status" tells me that there are "no instances running" and there also isn't a .pid file in ../tmp/pids

When I start resque-web I can see that the notifications are still queued and haven't been sent.

I can send the notifications manually by using "rake apn:sender", however "rake apn:senders" doesn't work.

Configuration: Rails 3.0.9, Passenger + nginx, deployed using Capistrano via git

The daemon works fine when I run the app locally on OS X.

Any help?

You wrote in README

sends them along to Apple over a single, persistent socket

but with Resque it creates new connection and new client object each time

D, [2013-11-21T13:05:55.474087 #17156] DEBUG -- : 2013-11-21 13:05:55 +0200: Connecting to gateway.push.apple.com:2195...
D, [2013-11-21T13:05:55.474160 #17156] DEBUG -- : 2013-11-21 13:05:55 +0200: #<APN::Client:0x007f3d00d888a0>
D, [2013-11-21T13:05:56.206347 #17156] DEBUG -- : 2013-11-21 13:05:56 +0200: Message sent.

...

D, [2013-11-21T13:06:01.213870 #17168] DEBUG -- : 2013-11-21 13:06:01 +0200: Connecting to gateway.push.apple.com:2195...
D, [2013-11-21T13:06:01.214269 #17168] DEBUG -- : 2013-11-21 13:06:01 +0200: #<APN::Client:0x007f3d00dba530>
D, [2013-11-21T13:06:01.807808 #17168] DEBUG -- : 2013-11-21 13:06:01 +0200: Message sent.

Maybe I'm doing something wrong, but can you explain this behavior?

The current payload size is set as a hard 255 byte limit, but iOS 8 has increased the payload size to 2KB. From Apple's documentation, it would appear it applies only to iOS 8, however. Relevant section:

We adjusted the limit in a fork but is just changing the limit globally desirable? Any preferences on a way to let them specify?