test(job-scheduler): add tests with drain and promote for job scheduler #2944
Conversation
roggervalf
changed the title
tests/test_job_scheduler.ts
Outdated
|
|
||
| await queue.promoteJobs(); | ||
|
|
||
| expect( |
There was a problem hiding this comment.
what about
const delayedCount = await queue.getDelayedCount();
const waitingCount = await queue.getWaitingCount();
expect(delayedCount).to.be.equal(1);
expect(waitingCount).to.be.equal(1);by separating each expectation will be easily to identify where something is failing
tests/test_job_scheduler.ts
Outdated
| expect( | ||
| (await queue.getDelayedCount()) == 1 || | ||
| (await queue.getWaitingCount()) == 1 || | ||
| (await queue.getActiveCount()) == 1, |
There was a problem hiding this comment.
as worker is having a no-operational processor, it could process a waiting job really fast and no be in active state, I recommend removing worker instance and this check
|
|
||
| await queue.upsertJobScheduler('scheduler-test', { | ||
| every: 50000, | ||
| immediately: true, |
There was a problem hiding this comment.
this option is not needed when using every option as it happens by default
tests/test_job_scheduler.ts
Outdated
| (await queue.getWaitingCount()) == 1 || | ||
| (await queue.getActiveCount()) == 1, | ||
| ).to.be.true; | ||
| await queue.removeJobScheduler('scheduler-test'); |
There was a problem hiding this comment.
may no be needed as afterEach hook is removing all keys for this queue
| every: 50000, | ||
| immediately: true, | ||
| }); | ||
| const worker = new Worker(queueName, async () => {}, { connection }); |
There was a problem hiding this comment.
maybe initialize worker after drain operation as worker could take the first iteration as quick as possible
| it('worker should start processing repeatable jobs after drain', async function () { | ||
| await queue.upsertJobScheduler('scheduler-test', { | ||
| every: 50000, | ||
| immediately: true, |
There was a problem hiding this comment.
immediately happens by default when using every option, no need to pass it
tests/test_job_scheduler.ts
Outdated
| immediately: true, | ||
| }); | ||
|
|
||
| expect(worker.isRunning()).to.be.true; |
There was a problem hiding this comment.
instead of this check, you are looking for checking that worker is able to process jobs from scheduler after drain operation, you can use a queueEvent instance to listen for completed state or from worker instance like in https://github.com/taskforcesh/bullmq/blob/master/tests/test_worker.ts#L1835-L1851
| await queue.promoteJobs(); | ||
|
|
||
| const waitingCount = await queue.getWaitingCount(); | ||
| expect(waitingCount).to.be.equal(1); |
There was a problem hiding this comment.
to check that at least the next repeatable was scheduled you may need to check the delayed count as well, so after this line you can add:
const delayedCount = await queue.getDeleayedCount();
expect(delayedCount).to.be.equal(1);| }); | ||
| }); | ||
|
|
||
| const job = await queue.add('test', { foo: 'bar' }); |
There was a problem hiding this comment.
this addition looks like it's not needed but you may need to check the completion of the scheduled job from scheduler
ref #247 #2405