BullMQ on AWS MemoryDB

[nullndr]

I'm having some trouble running BullMQ on MemoryDB.

I followed the redis cluster pattern, adding the prefix key on each of my queues.

I can see my queues in the cluster:

> keys *
1) "{queuePrefix1}:queue name 1:repeat:123456:7890"
2) "{queuePrefix2}:queue name 2:wait"
3) "{queuePrefix3}:queue name 3:id"
...

But I still get the following error:

ReplyError: CROSSSLOT Keys in request don't hash to the same slot
    at Scripts.moveToActive (./node_modules/bullmq/dist/cjs/classes/scripts.js:559:37)
    at runNextTicks (node:internal/process/task_queues:60:5)
    at listOnTimeout (node:internal/timers:540:9)
    at process.processTimers (node:internal/timers:514:7)
    at async Worker.moveToActive (./node_modules/bullmq/dist/cjs/classes/worker.js:236:55) {
  command: {
    name: 'evalsha',
    args: [
      ...
    ]
  }
}

This is just my assumption, but should I add the prefix key on the worker options too?

My connections are declared as the following:

const queueRedisConnection =
  process.env.NODE_ENV === "production"
    ? new Redis.Cluster([{ host: Env.get("REDIS_URL") }], {
        clusterRetryStrategy: () => null,
        enableReadyCheck: true,
        enableOfflineQueue: false,
        showFriendlyErrorStack: true,
        redisOptions: {
          tls: {
            checkServerIdentity: () => undefined,
          },
          username: "",
          password: "",
        },
      })
    : new Redis(Env.get("REDIS_URL"), queueRedisOptions);

const workerRedisConnection =
  process.env.NODE_ENV === "production"
    ? new Redis.Cluster([{ host: Env.get("REDIS_URL") }], {
        clusterRetryStrategy: () => null,
        enableReadyCheck: true,
        enableOfflineQueue: true,
        showFriendlyErrorStack: true,
        redisOptions: {
          tls: {
            checkServerIdentity: () => undefined,
          },
          username: "",
          password: "",
        },
      })
    : new Redis(Env.get("REDIS_URL"), workerRedisOptions);

[manast]

You need to add the prefix on the Queue and Worker instances.

[nullndr]

Yes, that's what I tried.
Would you be open to a PR to improve these docs?

[manast]

Sure. Anything that helps clarifying is valuable.