Leaked timeouts when shutting down

[cowboyd]

During test cases that involve starting and stopping a consumer many times, @platformatic/kafka leaks timers:

It appears that there is no teardown logic for this timer and it is never cleared.

kafka/src/clients/base/base.ts

Line 371 in 603f394

setTimeout(() => {

[mcollina]

How did you detect? Can upload a small test to reproduce?

[cowboyd]

It's in a test suite in a private repo. I'll see if I can't extract a simpler reproduction. In the mean time, here is the stack trace that was given to me by deno --trace-leaks

error: Leaks detected:
  - A timer was started in this test, but never completed. This is often caused by not calling `clearTimeout`. The operation was started here:
    at Object.queueUserTimer (ext:core/01_core.js:793:9)
    at setTimeout (ext:deno_web/02_timers.js:64:15)
    at Timeout.<computed> (ext:deno_node/internal/timers.mjs:76:7)
    at new Timeout (ext:deno_node/internal/timers.mjs:57:37)
    at setTimeout (node:timers:14:10)
    at .cache/deno/npm/registry.npmjs.org/@platformatic/kafka/1.6.0/dist/clients/base/base.js:245:21
    at Object.worked [as callback] (.cache/deno/npm/registry.npmjs.org/fastq/1.19.1/queue.js:222:14)
    at Connection.#onClose (.cache/deno/npm/registry.npmjs.org/@platformatic/kafka/1.6.0/dist/network/connection.js:427:22)
    at Socket.emit (ext:deno_node/_events.mjs:434:35)
    at node:net:1161:14

all producers and consumers have successfully resolved a forced close: close(true)