Lightweight SC4SNMP installation

SC4SNMP can be successfully installed in small environments with 2 CPUs and 4 GB of memory. One important thing to remember is that Splunk OpenTelemetry Collector for Kubernetes cannot be installed in such a small environment along with SC4SNMP. The other difference from normal installation is that the resources limits must be set for Kubernetes pods. See the example of values.yaml with the appropriate resources below:

splunk:
  enabled: true
  protocol: https
  host: ###SPLUNK_HOST###
  token: ###SPLUNK_TOKEN###
  insecureSSL: "false"
  port: "###SPLUNK_PORT###"
image:
  pullPolicy: "Always"
traps:
  replicaCount: 1
  resources:
     limits:
       cpu: 100m
       memory: 300Mi
     requests:
       cpu: 40m
       memory: 256Mi 
  communities:
    2c:
      - public
      - homelab
  #usernameSecrets:
  #  - sc4snmp-hlab-sha-aes
  #  - sc4snmp-hlab-sha-des

  #loadBalancerIP: The IP address in the metallb pool
  loadBalancerIP: ###X.X.X.X###
worker:
  # There are 3 types of workers 
  trap:
    # replicaCount: number of trap-worker pods which consumes trap tasks
    replicaCount: 1
    resources:
      limits:
        cpu: 100m
        memory: 300Mi
      requests:
        cpu: 40m
        memory: 150Mi
    #autoscaling: use it instead of replicaCount in order to make pods scalable by itself
    #autoscaling:
    #  enabled: true
    #  minReplicas: 2
    #  maxReplicas: 40
    #  targetCPUUtilizationPercentage: 80
  poller:
    # replicaCount: number of poller-worker pods which consumes polling tasks
    replicaCount: 2
    resources:
      limits:
        cpu: 200m
        memory: 600Mi
      requests:
        cpu: 60m
        memory: 260Mi
    #autoscaling: use it instead of replicaCount in order to make pods scalable by itself
    #autoscaling:
    #  enabled: true
    #  minReplicas: 2
    #  maxReplicas: 40
    #  targetCPUUtilizationPercentage: 80
  sender:
    # replicaCount: number of sender-worker pods which consumes sending tasks
    replicaCount: 1
    resources:
      limits:
        cpu: 100m
        memory: 350Mi
      requests:
        cpu: 30m
        memory: 250Mi
    # autoscaling: use it instead of replicaCount in order to make pods scalable by itself
    #autoscaling:
    #  enabled: true
    #  minReplicas: 2
    #  maxReplicas: 40
    #  targetCPUUtilizationPercentage: 80
  # udpConnectionTimeout: timeout in seconds for SNMP operations
  #udpConnectionTimeout: 5
  logLevel: "INFO"
scheduler:
  logLevel: "INFO"
  resources:
     limits:
       cpu: 40m
       memory: 260Mi
     requests:
       cpu: 20m
       memory: 180Mi
#  profiles: |
#    generic_switch:
#      frequency: 300
#      varBinds:
#        - ['SNMPv2-MIB', 'sysDescr']
#        - ['SNMPv2-MIB', 'sysName', 0]
#        - ['TCP-MIB', 'tcpActiveOpens']
#        - ['TCP-MIB', 'tcpAttemptFails']
#        - ['IF-MIB']
poller:
 # usernameSecrets:
 #   - sc4snmp-hlab-sha-aes
 #   - sc4snmp-hlab-sha-des
 # inventory: |
 #   address,port,version,community,secret,security_engine,walk_interval,profiles,smart_profiles,delete
 #   10.0.0.1,,3,,sc4snmp-hlab-sha-aes,,1800,,,
 #   10.0.0.199,,2c,public,,,3000,,,True
 #   10.0.0.100,,3,,sc4snmp-hlab-sha-des,,1800,,,
sim:
  # sim must be enabled if you want to use signalFx
  enabled: false
#  signalfxToken: BCwaJ_Ands4Xh7Nrg
#  signalfxRealm: us0
mongodb:
  pdb:
    create: true
  persistence:
    storageClass: "microk8s-hostpath"
  volumePermissions:
    enabled: true
inventory:
  resources:
     limits:
       cpu: 60m
       memory: 300Mi
     requests:
       cpu: 20m

The rest of the installation is the same as in online, or the offline installation.

Keep in mind that a lightweight instance of SC4SNMP won’t be able to poll from many devices and may experience delays if there is frequent polling.