K
Q

MySQL ram usages on default pod

4/21/2020

Im trying to run a basic MySQL 8 pod in Kubernetes. I did a basic deployment without any resource limits or whatsoever. What i do notice that the memory consumption is high. I have a nearly empty database (i think there are max 100 rows with basic data) and the pod is consuming 750M memory.

Is there anything that you can do about this?

apiVersion: apps/v1
kind: Deployment
metadata:
  name: db
  namespace: my-namespace
spec:
  replicas: 1
  selector:
    matchLabels:
      app: db
  strategy:
    type: Recreate
  template:
    metadata:
      labels:
        app: db
    spec:
      containers:
        - name: db
          image: mysql:8.0
          resources:
          env:
            - name: MYSQL_DATABASE
              value: mydb
            - name: MYSQL_USER
              value: myuser
            - name: MYSQL_PASSWORD
              valueFrom:
                secretKeyRef:
                  name: secret
                  key: "DATABASE_PASSWORD"
          ports:
            - containerPort: 3306
              name: transport
              protocol: TCP
          volumeMounts:
            - name: db
              mountPath: /var/lib/mysql
              subPath: mysql
      volumes:
        - name: db
          persistentVolumeClaim:
            claimName: db

Best Pim

-- Dirkos
docker
kubernetes
memory
mysql

Similar Questions

spring-cloud2 gateway ServerHttpRequest.getRemoteAddress in k8s very solw
Replace nginx ingresses with API gateway ambassador
Error on Websocket Implementation on Azure Kubernetes Cluster Not working
The kubernetes binary installation reported an error

3 Answers

4/23/2020

What i did to reduce the consumption mainly is disable the performance schema. Further i reduced the default connection limit to ensure that it does not reserve the memory for those who are not being used.

It will indeed mean that it can collapse if there are more users but i guess its a guess in on one of those

  • more cluster ram or failing to run it due to ram exhaustion
  • hit the max connection limit

Proper monitoring on all will do the trick

-- Dirkos
Source: StackOverflow
4/21/2020

In the future please add more information about your environment (local, kubeadm, minikube, cloud) and scenario. It would be easier to reproduce or troubleshoot.

I would not say that 750M RAM is high consumption. For test I've deployed MySQL 8.0 on my GKE cluster using HELM based on this chart. I've only changed default MySql image version to 8.0.

$ helm install sql stable/mysql

Pure instance of MySQL on my cluster without any limits, requests or data:

$ kubectl top pods
NAME                         CPU(cores)   MEMORY(bytes)
sql-mysql-6c9489d5b9-m8zmh   8m           376Mi

So if MySQL works with actual data it's normal to consume more resources.

I would say it's normal behaviour.

For general resources use in MySql you can check How MySQL Uses Memory.

However if you working on local environment with some limited resources you should specify Limits in your YAMLs.

-- PjoterS
Source: StackOverflow
4/21/2020

I've just ran a docker container with MySQL and it consumes ~400M of RAM with an empty database that is not even being queried by some application.

docker run -itd -e MYSQL_ROOT_PASSWORD=password mysql
docker stats
CONTAINER ID        NAME                CPU %               MEM USAGE / LIMIT     MEM %               NET I/O             BLOCK I/O           PIDS
a779ef705921        happy_lumiere       2.22%               373.4MiB / 1.943GiB   18.77%              1.05kB / 0B         0B / 0B             38

According to this assessment it seems to me that 750M is not that much.

You can check out these Best Practices for Configuring Optimal MySQL Memory Usage.

-- vkozyrev
Source: StackOverflow