Redis memory accumulates by job process return value

Using base use case inspired by quick start docs https://docs.bullmq.io/

const { Queue, Worker } = require('bullmq');

const someData = 'some data';

const queue = new Queue('queueName');

queue.add('index', {}, { removeOnComplete: true });

new Worker(queue.name, async job => {
  return someData;
});

Every time I run the code above, the job is added to the queue and processed. Redis memory increases (at least) linearly by someData size + some small amount.

I expect memory to decline after job is removed, which it supposedly is here - redis key bull:queueName:{id} not present upon completion.

Redis memory: redis-cli -> INFO MEMORY -> “used_memory”

This code equivalent in the old ‘bull’ library seems to work as expected.

[roggervalf]

drained event was saved continuously even when worker is idle, this was fixed in https://github.com/taskforcesh/bullmq/pull/1123, so I think this should be the main reason of this issue

[manast]

@sspread If you could provide a test case that demonstrates the issue I can look into it. As far as I can see, we XTRIM when posting events and in my tests it works as it should.