logicalparadox / apnagent Goto Github PK

On our production server, I received this error causing a server restart:

Error uncaughtException: DecoderStream 11 not implemented.

Error: DecoderStream 11 not implemented.
 at Decoder._nudge (/home/code/live/agentXserver/node_modules/apnagent/node_modules/lotus/lib/lotus/decoder.js:138:18)
 at Buffers.g (events.js:199:16)
 at Buffers.emit (events.js:107:17)
 at Buffers.push (/home/code/live/agentXserver/node_modules/apnagent/node_modules/lotus/node_modules/bufs/lib/bufs.js:73:10)
 at Decoder._write (/home/code/live/agentXserver/node_modules/apnagent/node_modules/lotus/lib/lotus/decoder.js:154:15)
 at doWrite (_stream_writable.js:301:12)
 at writeOrBuffer (_stream_writable.js:288:5)
 at Decoder.Writable.write (_stream_writable.js:217:11)
 at TLSSocket.<anonymous> (/home/code/live/agentXserver/node_modules/apnagent/lib/apnagent/agent/live.js:159:18)
 at TLSSocket.emit (events.js:107:17)
 at readableAddChunk (_stream_readable.js:163:16)
 at TLSSocket.Readable.push (_stream_readable.js:126:10)
 at TCP.onread (net.js:538:20)

How can I catch it?

Hi

I have a problem whit _header.js.

I have build an push server that works nice on my local developer env. But once I try to run it on server in DMZ ports opened are (2195,2196,80) if I run node agent/_header.js it doesn't connect and I dont get any error as well ?

this is from console on my Developer PC
D:\lpns>node agent/_header.js
apnagent [sandbox] gateway connected

if I do it the same way on the server I get nothing :(

PS C:\lpns\lpns> node agent/_header.js
PS C:\lpns\lpns>

What can be a issue ?

Hi,

I'm facing a strange problem where push notifications are working perfectly in the sandbox environment but not in production. We have verified that the .p12 certificate is working and can use the same certificate to send push notifications from another server application.

I'm using the node plugin for eclipse and stepping through the code and it appears that the notifications aren't being sent at all. The reason for this seems to be because the queue, to which the push tasks have been assigned, is paused - ie has its _paused field set to true.

However, no error is being thrown and as a result my node server thinks the push notifications have been sent, yet the devices never receive them.

It also appears that the exports.gatewayOptions function is being called repeatedly (I checked this to make sure it was picking up the correct host, in terms of sandbox and production) as a result of listOnTimeout() in timers.js, Agent._reconnect() in live.js, and Agent.connect() in live.js. For what it's worth, it is picking up the correct production URL, but the fact that this method is called repeatedly - every five seconds or so - made me wonder if there were somehow connection problems.

Can you help?

Thanks very much in advance.

Bob

The documentation refers to an 'error mitigation' section but I don't see any detail related to that section. Specifically, what to do when a user unsubscribes to notifications. My assumption is there is a message:error event emitted on the agent but I don't see the documentation describing it in any level of detail.

This library looks promising, thank you.

One question I had in mind is what's the recommended way to send push in bulk, i.e. sending the same message to 100K subscribers at the same time, preferably in one call ?

Cheers,
Michael

Messages should support the Safari/Mavericks push action dictionary.

apnagent.qualiancy.com is current down and has been for days. any eta on when it will be back up?

When I try to call apnagent.MockFeedback().unsub(token, new Date()), I got error "TypeError: value is out of bounds". What is the main reason for this error? Thanks.

Apnagent has default value for expires is 1d mean one day.

I am able to send notification to iOS device while device is online.

But, if i turn off internet connectivity & try to send notification it is not working.
I also try with setting agent.set('expires', '1d')

Below are the steps i am following :

1. Make device offline by turn off all internet connectivity
2. After that send push notification having expires value as 1 day
3. After two minutes i make device online by turning internet on
4. And i wait for 10 minute, still i am not getting any notification

I have this message on "createMessage"
"not cathed error:
[TypeError: undefined is not a function]"
but push sended and received well.
What is happened?

self.apnagent.createMessage()
.device(device)
.alert(message)
.badge(badge)
.sound('default')
.set('operation',operation)
.send(function (err) {
// handle apnagent custom errors
if (err) {
console.error('send push error:',err);
}
});
err - always "undefined"

Hi,
I'm trying send a MDM push notification to the APNS server. My connection is successful and sending of notification is successful. Unfortunately APNS is not delivering the notification to the device instead it is reported in feedback service. Below is the code I used
var apnAgent = require('apnagent'),
fs = require('fs');
var agent = new apnAgent.Agent();

var tokenBuff = new Buffer("", 'base64'); //device Token
var cert = fs.readFileSync('./certificates/newCrt.pem');
var key = fs.readFileSync('./certificates/newKey.pem');
var tokenHex = tokenBuff.toString('hex');
var settings = {
"device": tokenHex,
"pushMagic":"push magic token",
"auth": {
"cert": cert,
"key": key
}

};
agent.gateway = 'gateway.push.apple.com'
agent.set(settings.auth);

agent.connect(function (err) {
// gracefully handle auth problems
if ((err != null) && (err.name === 'GatewayAuthorizationError')) {
console.log('Authentication Error: %s', err.message);
process.exit(1);
}
// handle any other err (not likely)
else if (err) {
throw err;
}
else {
console.log('connect to APNS');
}
});
/**

• Listen for send errors
  • */

agent.on('message:error', function (err, msg) {
switch (err.name) {
case 'GatewayMessageError':
console.log('[emitted] gw notification error: %s', err.message);
if (err.code === 8) {
console.log(' > %s', msg.device().toString());

        }
        break;
    case 'MessageSerializationError':
        console.log('[emitted] serialization error: %s', err.message);
        break;
    default:
        console.log('[emitted] other error: %s', err.message);
        break;
}
//if (!err) {
//    console.log("Message is " + msg);
//}

});

var msg = agent.createMessage()
msg.device(tokenHex);
msg.set('mdm', settings.pushMagic);
msg.send(function (err) {
if (err) console.log('[cb] serialization error: %s', err.message);

});
The resultant Payload that is sent to server is using apnagent is
{"deviceToken":{"type":"Buffer","data":[]},"expiration":0,"identifier":0,"payload":{"mdm":"","aps":{}}}

Please help.

Thanks and regards
Venkatesh

It seems there's another bug after fixing #14 and #15 which is causing our agent to stop processing its queue.

Order of events:

1. 18:23:22 - Last notification sent
2. Queue begins piling up
3. 18:24:16 - Cache cleaned out (30s TTL)
4. 18:26:00 - Invalid token error received
5. 18:26:03 - Gateway successfully reconnects
6. Queue continues piling up. No more notifications have been sent since event 1 above.

I don't have any ideas as to why this is happening yet. Let me know if I can provide any more info that might be helpful.

How can I increment the badge count by one for every push notification sent?

How can i set up the apns-collapse-id as seen in the Apple documentation ?

We're having issues with agents not reconnecting sometimes after a message error. Here are the relevant log entries:

2014-08-04T20:03:58.212822+00:00 app[web.1]: apnagent:agent-live:web.1.14 [6ms] (gateway) data: 6 bytes
2014-08-04T20:03:58.214107+00:00 app[web.1]: apnagent:agent-live:web.1.14 [2ms] (gateway) disconnected - gateway.push.apple.com:2195
2014-08-04T20:03:58.214248+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (gateway) end
2014-08-04T20:03:58.214862+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (encoder) write: 28273
2014-08-04T20:03:58.215484+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (queue) pause: not connected
2014-08-04T20:03:58.198696+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (gateway) write: 28267
2014-08-04T20:03:58.199419+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (cache) push: 28267
2014-08-04T20:03:58.200158+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (encoder) write: 28268
2014-08-04T20:03:58.200419+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (gateway) write: 28268
2014-08-04T20:03:58.200600+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (cache) push: 28268
2014-08-04T20:03:58.200749+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (encoder) write: 28269
2014-08-04T20:03:58.201017+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (gateway) write: 28269
2014-08-04T20:03:58.201155+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (cache) push: 28269
2014-08-04T20:03:58.201307+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (encoder) write: 28270
2014-08-04T20:03:58.201578+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (gateway) write: 28270
2014-08-04T20:03:58.201718+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (cache) push: 28270
2014-08-04T20:03:58.201865+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (encoder) write: 28271
2014-08-04T20:03:58.202116+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (gateway) write: 28271
2014-08-04T20:03:58.202255+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (cache) push: 28271
2014-08-04T20:03:58.204973+00:00 app[web.1]: apnagent:agent-live:web.1.14 [2ms] (encoder) write: 28272
2014-08-04T20:03:58.205635+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (gateway) write: 28272
2014-08-04T20:03:58.206080+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (cache) push: 28272
2014-08-04T20:04:01.207280+00:00 app[web.1]: apnagent:agent-live:web.1.14 [2992ms] (cache) since: 28263
2014-08-04T20:04:01.217101+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28271
2014-08-04T20:04:01.215785+00:00 app[web.1]: apnagent:agent-live:web.1.14 [8ms] (queue) push: 28264
2014-08-04T20:04:01.215850+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28265
2014-08-04T20:04:01.215921+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28266
2014-08-04T20:04:01.215987+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28267
2014-08-04T20:04:01.216170+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (queue) push: 28268
2014-08-04T20:04:01.216226+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28269
2014-08-04T20:04:01.216590+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (queue) push: 28270
2014-08-04T20:04:01.217914+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (queue) push: 28272
2014-08-04T20:04:01.217969+00:00 app[web.1]: apnagent:agent-live:web.1.14 [0ms] (gateway) reconnecting
2014-08-04T20:04:01.218187+00:00 app[web.1]: apnagent:agent-live:web.1.14 [1ms] (gateway) connecting - gateway.push.apple.com:2195

You can see that the gateway receives 6 bytes, then closes the gateway as it should. Next we see the queue pause, which is also expected. Then, two strange things happen. First, we see that after the queue has been paused, the gateway is still being written to, which shouldn't happen. The next strange thing is that although the reconnect attempt happens 3 seconds later, it never finishes connecting or throws an error. That's the last line in our log for that worker after it continued to run for another hour or so. I'm going to experiment with setTimeout and the timeout event on the tls socket to see if that event gets called in this case. Very strange stuff...

Edit: As you can see I added some additional logging, which you can find on my fork if it's of interest.

I'm seeing consistent memory leaks in our apnagent process. We're running two clusters of 4 express servers / apnagents and after a few hours they both exceed Heroku's memory cap of 512 MB. I can confirm that during the same period, the cache and queues are staying around the same size. I'm running with a 30s cache ttl.

I'm fairly confident that the leak isn't in our application code, since it's really a dead-simple express app that only interacts with apnagent. That said, I'm hoping to reproduce and trace the leak on my dev machine soon.

I'm not retrieving any error messages, but after the server has been running for some time, notifications no longer shows up on the devices ?

When i restart the server everything works again :S

I'm running production mode, any thoughts ?

Will apnagent be compatible after this update: https://developer.apple.com/news/?id=10222014a ?

As it currently stands my professional interests have moved away from APN so I don't have as much time to maintain this as I used. Given its production value and importance I am open to allowing others to maintain this on a day-to-day basis.

If you or your company have committed to using apnagent and are interested in keeping this project moving, open a PR for an outstanding issue or new feature and mention your interest in maintaining so we can chat. Alternatively, feel free to email me through my GitHub profile. I will also provide access to any dependency which I have written for this module so you can keep things moving along.

Hi,
I was earlier able to send notifications using the below code.

agent
.set('cert file', join(__dirname, 'certfile.pem'))
.set('key file', join(__dirname, 'key.pem'))
.enable('sandbox');

agent
    .set('expires', '1d')
    .set('reconnect delay', '1s')
    .set('passphrase', 'password')
    .set('cache ttl', '30m');

agent.on('message:error', function (err, msg) {
    if(err) {
        console.log(err);
    }else{
        console.log(msg);
    }
});

agent.connect(function (err) {
    if (err && err.name === 'GatewayAuthorizationError') {
        console.log('Authentication Error: %s', err.message);
        process.exit(1);
    }

    // handle any other err (not likely)
    else if (err) {
        throw err;
    }

    // it worked!
    var env = agent.enabled('sandbox')
        ? 'sandbox'
        : 'production';

    console.log('apnagent [%s] gateway connected', env);
});

But suddenly, the agent has stopped connecting. eg when i console agent it shows :

{ _drip: { delimeter: ':', wildcard: true },
data: :11930 - connected: false,
data: :11930 - settings:
data: :11930 - { 'cache ttl': '30m',
data: :11930 - codec: 'enhanced',
data: : 11930 - }

It is not throwing an error neither is it sending messages. Any leads would be great. Thanks.

I'm developing a website , that should send push notifications to iOS devices. On my localhost apnagent works fine, but when I git push it to Openshift cloud I get the following errors in the terminal:

remote: Waiting for application port (8080) become available ...
remote: Application 'apnagent' failed to start (port 8080 not available)
remote: -------------------------
remote: Git Post-Receive Result: failure
remote: Activation status: failure
remote: Activation failed for the following gears:
remote: 55c3344c89f5cf720xxxxxxx (Error activating gear: CLIENT_ERROR: Failed to execute: 'control start' for /var/lib/openshift/55c3344c89f5cf720xxxxxxx/nodejs
remote: #<IO:0x00000000f1c278>
remote: #<IO:0x00000000f1c200>
remote: )
remote: Deployment completed with status: failure
remote: postreceive failed
To ssh://[email protected]/~/git/apnagent.git/
   bbcf2a3..87exxxx  master -> master 

My package.json looks like this:

{
  "private": true,
  "name": "apnagent",
  "version": "0.0.0",
  "dependencies": {
    "apnagent": "1.0.x"
  },
  "engines": {
    "node": ">=0.10.22",
    "npm": ">=1.3.14"
  },
  "scripts": {
    "start": "node agent/_header.js"
  },
  "main": "agent/_header.js"
}

I think i should use somehow process.env.OPENSHIFT_NODEJS_PORT as port on the server, or something like this. I'm sure, I've done some configuration mistake since on my local machine everything works fine.

We're seeing intermittent SSL errors that produce the following log output:

2014-08-01T18:31:59.274859+00:00 app[web.2]: apnagent:agent-live [278ms] (gateway) error: 139718328477472:error:14094416:SSL routines:SSL3_READ_BYTES:sslv3 alert certificate unknown:../deps/openssl/openssl/ssl/s3_pkt.c:1275:SSL alert number 46
2014-08-01T18:31:59.275014+00:00 app[web.2]: Gateway error [Error: 139718328477472:error:14094416:SSL routines:SSL3_READ_BYTES:sslv3 alert certificate unknown:../deps/openssl/openssl/ssl/s3_pkt.c:1275:SSL alert number 46

No idea why we're seeing the error, given that the certs work fine 99% of the time. The main issue though is that the agent's queue is never restarted and/or a connection to the gateway is not made, so our application eventually runs out of memory from the queue backing up.