Cron just stops - Possible Memory Leak

[StevenDStanton]

I have a Cron job that just randomly stops. I started it at around 7:30PM and by 8:10 PM it just stopped.

I was logging output to my console and it just died.

 new CronJob('*/5 * * * * *', function(){
       console.log("***************************************************");
       console.log("Cron Start: " + Date());
       var https = require('https');
       var options = {
            host: 'api.twilio.com',
            port: 443,
            path: '/2010-04-01/Accounts/' + sid + '/Queues/'+ config.twilio.queue.sid  + '/Members.json',
            method: 'GET',
            auth: sid + ":" + auth,
            agent: false
       };
       var req = https.request(options, function(rex) {
         rex.setEncoding('utf8');
         rex.on('data', function (chunk) {
           try{
                  console.log("Cron Work : " + Date());
                 obj = JSON.parse(chunk);
                 io.sockets.emit('queue', {msg:obj.total.toString()});
                 io.sockets.emit('list', obj);
                 smsModel.find({agent: null}, function(err, results){
                          io.sockets.emit('chat', results);
                  console.log("Cron Data : " + Date());
              //console.log(results);
        });

      }catch(err){
        console.log("Code 1: " + err.message);
      } 
    });
  });
  req.end();
  console.log("Cron End  : " + Date());
  console.log("***************************************************");
}, null, true, "America/Los_Angeles");

It never logged any errors from my try catch blocks.

[StevenDStanton]

I tested further setting different cron times. I did it at 5, 10, and 20 second intervals and it seemed to have no effect on when it would fail.

I also monitored memory closely and while memory usage kept going up more and more while it was running until the crash it never maxed memory out.

Sometimes it would fail after 30 minutes, sometimes it would fail after 8 hours. It just seems to randomly and silently fail. Even when in a try catch block, in debugging mode, it just stops processing and goes blank with no errors reported.

all other portions of the server are working when this failure happens

[nathschmidt]

@StevenDStanton What version are you using? I've just had a similar issue with version 1.0.6, though there are some commits in the 1.0.8 release that fixes a similar issue #141

[StevenDStanton]

I am using the latest version which is on NPM. According to my package.json file it is 1.0.8

[ncb000gt]

@StevenDStanton I published a new version after this issue was reported. Are you still seeing the hanging behavior?

[StevenDStanton]

I was experiencing it with version 1.0.8 3 days ago.

I had to stop using the software and move on to querying the server from the client side to get my updates to move forward with the project I am working on.

If you have released an update since 1.0.8 I can download it and test

[jBachalo]

Has this issue been resolved? Really important to know as a current project will be kiosk based and running 24-7.

[h0x91b]

I am fixed it, in our production servers works ok
On Sun, Apr 12, 2015 at 04:02 jBachalo notifications@github.com wrote:

Has this issue been resolved? Really important to know as a current
project will be kiosk based and running 24-7.

—
Reply to this email directly or view it on GitHub
https://github.com/ncb000gt/node-cron/issues/149#issuecomment-91956985.

[ncb000gt]

@jBachalo the "stopping" bit was unrelated to a memory leak. That should be fixed in the latest version (although it sounds like @StevenDStanton is still having issues?).

I haven't seen or been able to reproduce a memory leak in the module locally, but that doesn't mean there isn't one. There is always a possibility that bugs exist. I'm not sure exactly what testing was done to prove a memory leak, but if the sample above is an example of what was tested then I'd recommend removing any excess code to rule out that as the memory leak culprit and seeing if the issue still exists.

[StevenDStanton]

I stopped using it. We are moving forward quickly on the project I was using it on and we had to get a replacement. I plan to come back to it later on another project to see if it happens again

[ncb000gt]

Ok. I'm going to close this for now. Can revisit if you do see a memory leak.

[mek-omkar]

i am facing similar problem, with cron 1.1.0 & nodev6.1.0. i configured every second job & not configured timezone

[semy]

Same problem.

cron 1.1.0 & nodev5.4.1
every second job with timezone

[semy]

@StevenDStanton What's your replacement of cron? Need a quick solution that works... Thanks :-)

[StevenDStanton]

@semy I am actually just using two different methods.

Where it is not critical that my calls be exact to the second I am using javascripts setInterval.

It has proven to be surprising robust and accurate. Currently it runs my main loop which pushes out status updates to around 50 agents every 10 seconds.

For to the second precision I am using actual cron on my linux server. That is simple to set up with cron tab and cURL.