question

Dharun avatar image
Dharun asked Erick Ramirez answered

DSE 6.8.24 does not reflect the memory and cores in Spark UI

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: dse
  labels:
    app: dse
spec:
  serviceName: dse
  replicas: 3
  selector:
    matchLabels:
      app: dse
  template:
    metadata:
      labels:
        app: dse
    spec:
      terminationGracePeriodSeconds: 1800
      containers:
      - name: dse
        image: datastax/dse-server:6.8.24
        args: ["-k"]
        imagePullPolicy: Always
        ports:
        - containerPort: 7000
          name: intra-node
        - containerPort: 7001
          name: tls-intra-node
        - containerPort: 7199
          name: jmx
        - containerPort: 9042
          name: cql
        - containerPort: 9142
          name: cql-ssl
        - containerPort: 7077
          name: spark
        - containerPort: 4040
          name: sparkui
        resources:
          limits:
            cpu: "3000m"
            memory: 6Gi
          requests:
            cpu: "3000m"
            memory: 6Gi
        securityContext:
          capabilities:
            add:
              - IPC_LOCK
        lifecycle:
          preStop:
            exec:
              command: 
              - /bin/sh
              - -c
              - nodetool drain
        env:
          - name: JVM_EXTRA_OPTS
            value: -Xmx7g -Xms5g
          - name: SEEDS
            value: dse-0.dse.bnc.svc.cluster.local
          - name: CLUSTER_NAME
            value: "bnc-cluster"
          - name: DC
            value: "DC1"
          - name: RACK
            value: "Rack1"
          - name: "DS_LICENSE"
            value: "accept"
          - name: NUM_TOKENS
            value: '2'
        readinessProbe:
          exec:
            command:
            - /bin/bash
            - -c
            - cqlsh -e "desc full schema"
          initialDelaySeconds: 120
          timeoutSeconds: 5
          periodSeconds: 10


I have give 3 core and 6GB of ram, however in the spark UI i see
1571Mb of ram

I see it works fine with 6.8.20 version, do i need to add any additional parameters while deploying ?screenshot-2022-07-05-at-114137-am.png

dockeranalytics
2 comments
10 |1000

Up to 8 attachments (including images) can be used with a maximum of 1.0 MiB each and 10.0 MiB total.

smadhavan avatar image smadhavan ♦ commented ·

@Dharun, one of the colleagues here rightly pointed it out that you're supplying only `6BG` of RAM, but are requesting for `7GB` of (max) heap memory and this setup would eventually blow up. Did the same setup work with DSE `6.8.20` really?

0 Likes 0 ·
Dharun avatar image Dharun smadhavan ♦ commented ·
@smadhavan Apologies, I copied the testing deployment yaml file.

I tried with 8GB not 6GB


0 Likes 0 ·

1 Answer

Erick Ramirez avatar image
Erick Ramirez answered

That indicates to me that the server you're trying to run DSE does not have a lot of memory so only 1.5GB of RAM is getting allocated to Spark.

You cannot allocate more memory to Spark than what is available on the server after the majority has been allocated to DSE.

You will need to log a ticket with DataStax Support to get assistance with your environment. Cheers!

Share
10 |1000

Up to 8 attachments (including images) can be used with a maximum of 1.0 MiB each and 10.0 MiB total.