Memory limit when running in k8s

[lethehoa]

Currently, I install dragonflydb in k8s. Here is my manifest file:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: dragonflydb
  labels:
    app: dragonflydb
spec:
  selector:
    matchLabels:
      app: dragonflydb
  template:
    metadata:
      labels:
        app: dragonflydb
    spec:
      containers:
      - name: dragonflydb
        image: docker.dragonflydb.io/dragonflydb/dragonfly:latest
        resources:
          limits:
            memory: 1024Mi
            cpu: 200m
          requests:
            memory: 1024Mi
            cpu: 200m
        args:
          - --maxmemory=512M
          - --cache_mode=true
        ports:
        - containerPort: 6379

But I have the problem below.

I20231202 05:28:49.835762 1 init.cc:70] dragonfly running in opt mode. I20231202 05:28:49.836225 1 dfly_main.cc:800] Starting dragonfly df-v1.12.1-053a33d24df939d1c35f3b6f233a0c57ac185184 W20231202 05:28:49.836840 1 dfly_main.cc:839] SWAP is enabled. Consider disabling it when running Dragonfly. I20231202 05:28:49.836941 1 dfly_main.cc:863] Max memory limit is: 512.00MiB W20231202 05:28:49.836977 1 dfly_main.cc:504] Kernel is older than 5.10, switching to epoll engine. I20231202 05:28:49.994577 1 proactor_pool.cc:147] Running 24 io threads E20231202 05:28:49.994702 1 dfly_main.cc:336] There are 24 threads, so 6.00GiB are required. Exiting...

[lethehoa]

#1453
I solved my problem by set suitable value for arg: proactor_threads like the discussion above.