Ability not to start job after delay

[niko-lay]

Hello
I'm looking for feature to not start job processing if some delay elapsed time since the addition to queue. I assumed that parameter Job#delay is designed for this porposes, but it does not seem like this.

Here is minimal example

const Queue = require('bull');
const m = require('moment');
const conf = require('./lib/config.js');


const JOB_PROCESS_TIME = 15 * 1000; // simulate long job
const JOB_T = 3* 1000;
const JOB_D = 1 * 1000;

// process 1 job per 30 sec
var testQ = new Queue('testQ', {limiter: {max: 1, duration: 30 * 1000}, redis: conf.redisConnection });
testQ.empty()
    .then(()=>{
        console.log('testQ is empty now');

        for (i = 0; i< 3; i++){
            testQ.add({data: 'test-test-test', dateStrAdded: m().format(), date: new m()},
                      {delay: JOB_D, timeout: JOB_T})
                    .then((job)=>{
                        console.log('job added: ' + job.id);
                    })
                    .catch(console.error);
        }
    });

testQ.process(function(job, done){
    console.log(`Request job process: ${job.id} ${m().format()}`);
    setTimeout(()=>{
        var diff = m() - m(job.data.date);
        console.log("job processed : " + job.id + ' ' + m().format() + ' diff: '  + diff);

        done(null, {data: 'someData'})
    }, JOB_PROCESS_TIME);
})


testQ
.on('completed', function(job, result){
    //console.log("completed: " + job.id + ' ' + JSON.stringify(result));
    console.log("completed: " + JSON.stringify(job));
})
.on('failed', function(job, err){
    var diff = m() - m(job.data.date);
    console.log("failed: " + job.id + ' ' + m().format() + ' diff: '  + diff);

})

and output

testQ is empty now
job added: 19
job added: 20
job added: 21 <<<<<<<<<<< all added at once
Request job process: 19 2018-09-19T17:33:40+04:00
failed: 19 2018-09-19T17:33:45+04:00 diff: 6792 <<<<<<<<<<< now I would like all jobs failed, but only the first did it
job processed : 19 2018-09-19T17:33:55+04:00 diff: 16463
Request job process: 20 2018-09-19T17:34:10+04:00
failed: 20 2018-09-19T17:34:15+04:00 diff: 36952
job processed : 20 2018-09-19T17:34:25+04:00 diff: 46627
Request job process: 21 2018-09-19T17:34:40+04:00
failed: 21 2018-09-19T17:34:45+04:00 diff: 67115
job processed : 21 2018-09-19T17:34:55+04:00 diff: 76790

bull version v.3.4.8

Any ability to not start/remove/fail job after a certain time after its addition, not start processing?

[manast]

you will need to rephrase what are your expected behaviour, since I am not able to comprehend it based on your description.

[niko-lay]

I need something like TTL for single job or for all queue.

[manast]

Something like this? #479

[niko-lay]

yes. I seems this not yet implemented in bull?

[manast]

The issue is still open so no. I will close then this one as a duplicated.