Deploy the Tenant Service

From Genesys Documentation
Jump to: navigation, search

Learn how to deploy the Tenant Service.

Related documentation:
Important
Make sure to review Before you begin for the full list of prerequisites required to deploy the Tenant Service.

Deployment scenarios

More than one deployment scenario is supported, including single region, redundant, and multi-region deployment as well as multi-Tenant deployment.

Single region/location/cluster

You deploy Tenant resources in a single Kubernetes cluster within the same or separate namespace (project) with the Voice platform. If shared resources are being deployed across all Tenants, they must also be added to the same target namespace.

The Tenant deployment process creates resources using a release name parameter, specified when executing the Helm deployment step. When installed in a single namespace, you must make sure that the release name value is distinct across all Tenants and other deployments.

For example, you might specify the Helm release name in the format t<short_id>. Optionally, if you want the Tenant service name to match other Voice services, you can prefix the Tenant name with voice- in the Helm release name. So, in this example, you would specify the release name as voice-t<short_id> during Helm deployment. The value for <short_id> is the last four characters of the Tenant UUID that you configure in the values.yaml file. For more information about the identification parameters for the Tenant service, see Identification.

If you plan to use Prometheus monitoring or Fluent Bit logging framework for Tenant, you must execute the tenant-monitor module, as described in tenant-monitor. The module enables the following features:

  • Prometheus PodMonitor definition for all tenant pods.
  • Common Fluent Bit framework configuration for all tenant pods.

Single Tenant: Basic deployment

Single-node deployment requires a single override file and one "tenant" module to deploy, with reference implementation described at Single service at one location.

Mandatory parameters for basic installation are:

  • tenant uuid (v4)
  • tenant nickname (becomes a Helm release name)
  • all backend parameters (along with all secrets that may be required based on these parameters)

To increase the number of nodes, adjust the node count parameter. For more information, see Scalability and redundancy parameters.

Upgrade

An upgrade can be performed by re-running deployment with the same mandatory parameters and adjusted version of Tenant image(s) and Helm charts. The upgrade is performed automatically, one node at a time (if node count is > 1). The update of the Tenant configuration may happen automatically when the primary node is upgraded. Genesys recommends performing a backup of the backend database for a Tenant Service before upgrading.

Multiple Tenants at one location: Basic deployment

You can deploy additional Tenants at the same location using the following guidelines:

  • Each Tenant Service must have a unique tenant uuid, shortid, and nickname.
  • Each Tenant Service is deployed or upgraded and adjusted independently.

Multiple regions/locations/clusters: Basic deployment

In multi-regional/multi-location deployments, one region/location is considered "master" (from the Tenant perspective) and includes the database backend with write capabilities. Other regions/locations have replicas of the database backend in read-only mode. A Tenant Service at each location may be deployed to have one of its nodes running as master (write access to provisioning data through the config API) or have all its nodes running only as replicas (read access to configuration).

Multi-regional deployments must be performed using the following steps (with prerequisites already satisfied at each region/location):

  • If required, deploy the tenant-monitor module at a location planned as a Master Tenant node.
  • Complete the basic deployment of a Tenant Service in the Master region, including specification of DR parameters for the Master, as per Scalability and redundancy parameters.
  • Complete the deployment of the database backend with replica of the Master database at location(s) where the replica Tenant nodes are expected to run, including provisioning of access keys/secrets to access the local replica.
  • If required, deploy the tenant-monitor module at location(s) where replicas are expected to run.
  • Complete the basic Tenant deployment for additional region(s) and specify DR parameters for the Master region (see Scalability and redundancy parameters).

The same customization scenarios described for Tenant nodes can be applied for each location independently.  

Upgrade

An upgrade can be performed by re-running the deployment with the same mandatory parameters and adjusted version of the Tenant image(s) and Helm charts for every location. Update of the tenant configuration may happen automatically when the Master node in the Master region is upgraded. Genesys recommends performing a backup of the backend database for a Tenant Service before you upgrade the Master region.

Deploy the service

This section provides reference commands with key parameters that are required to complete each deployment step.

On this page, the tenant-values.yaml file refers to the values.yaml file in the Tenant Helm chart. Likewise, the tenant-monitor-values.yaml file refers to the values.yaml file in the Tenant Monitor Helm chart.

Prerequisite

Before proceeding with the Tenant Service deployment, ensure you have completed procedures in the Configure security section of this guide.


Location-specific deployment steps

tenant-monitor

Monitoring/logging shared configuration and infrastructure deployment:
helm upgrade --install --force --wait --timeout 600s -n voice  tenant-monitor https://<jfrog artifactory/helm location>/tenant-monitor-$TENANT_MANIFEST_VERSION.tgz  --username "$JFROG_USER" --password "$JFROG_PASSWORD"

To enable Prometheus monitoring, you can use the following overrides with tenant-monitor. Use the following changes in the tenant-monitor-values.yaml file to implement the changes:

prometheus:
  podMonitor:
    create: "true"

To enable Fluent Bit to send additional logs to stdout in json format (for selected Tenant functions, such as configuration audit) and/or raw format (such as from internal applications such as StatServer and URS), modify the following changes to upgrade tenant-monitor. Use the following changes in the tenant-monitor-values.yaml file to implement the changes.

fluent:
  enable: "true"
  rawlogs:
    stdout:
      enable: "true"
  jsonlogs:
    stdout:
      enable: "true"

To enable RWX Persistent Volume Claim (PVC) in tenant-monitor to store Tenant logs shared across all Tenant pods, make the following modifications to override values in the tenant-monitor-values.yaml file:

tenant:
  logging:
    volume:
      enabled: "true"
      createSC: "false"
      createpvClaim: "true"
      logClaim: "tenant-logs-pvc"
      logClaimSize: "5Gi"
      logStorageClass: "<storage class>"
      Storageprovisioner: "TBD OC provisioner"
      parameters: {}

RWX PV is disabled by default; no overrides are required in the tenant-monitor-values.yaml file to disable it.


Service-specific deployment steps: Single service at one location

A PostgreSQL database must be available for the Tenant Service before you begin the service deployment. For more information about the database requirements, see Third-party prerequisites. In addition, after the PostgreSQL database is deployed and before you deploy the Tenant Service, you must configure secrets that contain values for certain PostgreSQL database parameters. To configure the secrets, see Service-specific secrets.

Use the following template if you are deploying with the tenant Helm chart. A single-service deployment can be implemented with the following sample parameters in the tenant-values.yaml file:

##UUID 4 format ( Set a new UUID for new tenant deployment)
tenantid: <tenant-uuid>
 
 
serviceAccount:
    create: true
 
images:
  imagePullSecrets: mycred
  registry: <docker-repo> 
  pullPolicy: Always
  tenant:
    tag: <version of tenant service>
 
  pgdbInit:
    tag: <version of tenant service>
   
  rcsInit:
    tag: <version of roles service batch image>
    enable: "true"
 
  pulseInit:
    tag: <version of tenant service>
    enable: "true"
    pulseMode: "setup"
 
   
 
tenant:
  general:
    upstreamServices: voice-sipfe:9101,voice-config:9100,ixn-server-{{ $.Values.tenantid }}:7120,ixn-vqnode-{{ $.Values.tenantid }}:7122"
  pgdb:
    dbhost: "/opt/genesys/dbserver/dbserver"
    dbuser: "/opt/genesys/dbuser/dbuser"
    dbname: "/opt/genesys/dbname/dbname"
  securityContext:
    fsGroup: 0        
 
  logging
 ....
  volumes: 
    logPvc:
        enabled: "true"
        logClaimSize: "5Gi"
        accessModes: "ReadWriteOnce"
        logStorageClass: "<storage class>"   #Replace the storage class that's relevant to the Openshift Cluster

    mounts:
      log:
        - name: log
          mountPath: /opt/genesys/logs/volume
        - name: log
          mountPath: /logs
  secrets:
    pgdb:
      pwd:            
        secretName: "/opt/genesys/dbpassword/dbpassword"
        secretKey: "dbpassword"
      volumes: | 
        - name: dbpassword
          secret:
            secretName: dbpassword
        - name: dbserver
          secret:
            secretName: dbserver
        - name: dbname
          secret:
            secretName: dbname
        - name: dbuser
          secret:
            secretName: dbuser                        
      mounts:
        - name: dbpassword
          readOnly: true
          mountPath: "/opt/genesys/dbpassword"
        - name: dbserver
          readOnly: true
          mountPath: "/opt/genesys/dbserver" 
        - name: dbname
          readOnly: true
          mountPath: "/opt/genesys/dbname"  
        - name: dbuser
          readOnly: true
          mountPath: "/opt/genesys/dbuser"                          
 
    consul:
      acl:
        secretName: "/opt/genesys/consul-shared-secret/consul-consul-voice-token"
      volumes:
        - name: consul-shared-secret
          secret:
            secretName: consul-voice-token              
      mounts:
        - name: consul-shared-secret
          readOnly: true
          mountPath: "/opt/genesys/consul-shared-secret"     
    redis:
      configPwd:
        secretName: "/opt/genesys/redis-config-secret/redis-config-state"
        volumes:
          - name: redis-config-secret
            secret:
              secretName: redis-config-token              
        mounts:
          - name: redis-config-secret
            readOnly: true
            mountPath: "/opt/genesys/redis-config-secret" 
         
      streamPwd:          
        secretName: "/opt/genesys/redis-tenant-secret/redis-tenant-stream"
        volumes:
          - name: redis-tenant-secret
            secret:
              secretName: redis-tenant-token              
        mounts:
          - name: redis-tenant-secret
            readOnly: true
            mountPath: "/opt/genesys/redis-tenant-secret"        
 
    kafka:
      pwd:
         secretName: "/opt/genesys/kafka-secrets/kafka-secrets"
      volumes:  
        - name: kafka-secrets
          secret:
            secretName: kafka-secrets-token
      mounts:
        - name: kafka-secrets
          mountPath: "/opt/genesys/kafka-secrets"
    gws:
      user:
        secretName: "/opt/genesys/gauth-client-id/clientid"
      pwd:
        secretName: "/opt/genesys/gauth-client-token/clientsecret"
      volumes:
        - name: gauth-client-id
          secret:
            secretName: gauthclientid    
        - name: gauth-client-token
          secret:
            secretName: gauthclientsecret     
      mounts:          
        - name: gauth-client-id
          readOnly: true
          mountPath: "/opt/genesys/gauth-client-id"
        - name: gauth-client-token
          readOnly: true
          mountPath: "/opt/genesys/gauth-client-token"
 
 
 
  redis:
    isCluster: true

In addition, use the following deployment command:

helm upgrade --install --force --wait --timeout 600s -n voice -f ./tenant-node-values.yaml t<shortid> \
https://<jfrog artifactory/helm location>/tenant-<helm version>.tgz \
--username "$JFROG_USER" --password "$JFROG_PASSWORD"

The preceding deployment will create a Tenant with the password of the service account set up explicitly and without enabling GWS integration. See Samples and references for values that allow you to reset the Tenant password upon deployment using a pre-generated value from the secret and to enable automated GWS integration.

Samples and references

Enabling a service admin password (the secret should be created as described in the Service account password section):

...
tenant:
  serviceuser: "default"
  svcpwdSecretName: "/opt/genesys/service-user-account/svcpassword"
  ...
  volumes: |
    - name: service-user-account
      secret:
           secretName: svcuseraccount
    ...
  volumeMounts: |
    - name: service-user-account
      readOnly: true
      mountPath: "/opt/genesys/service-user-account"   
    ....
  initVolumeMounts: |        
    - name: service-user-account
      readOnly: true
      mountPath: "/opt/genesys/service-user-account"   
    ....

To enable stdout log output for all Tenant components, make the following modifications to override values in the tenant-values.yaml file.

images
fbregistry: fluent/fluent-bit
...
  fluentBit:
    enable: "true"
    name: json-sidecar
    tag: 1.8.x
 
  fluentBitUrs:
    enable: "true"
    name: stdouturs-sidecar
    tag: 1.8.x
 
  fluentBitSs:
    enable: "true" 
    name: stdoutss-sidecar
    tag: 1.8.x
 
  fluentBitOcs:
    enable: "true"
    name: stdoutocs-sidecar
    tag: 1.8.x
 
  fluentBitCs:
    enable: "true"
    name: stdoutcs-sidecar
    tag: 1.8.x
 
tenant:
 ...
  logging:
    volumes:
      log:
        - name: log
      jsonLog:
        - name: fluent-logs   
          emptyDir: {}
      stdoutUrsLog:
        - name: fluenturs-logs
          emptyDir: {}
      stdoutOcsLog:
        - name: fluentocs-logs
          emptyDir: {}
      stdoutSsLog:
        - name: fluentss-logs
          emptyDir: {}
      stdoutCsLog:
        - name: fluentcs-logs
          emptyDir: {}         
      fluentBconfigmap:
        - configMap:
            defaultMode: 420       
            name: tenants-fluent-bit-config
          name: tenants-fluent-bit-config         
      fluentBconfigmapCs:
        - configMap:
            defaultMode: 420       
            name: tenants-fluent-bit-config-cs
          name: tenants-fluent-bit-config-cs
      fluentBconfigmapSs:
        - configMap:
            defaultMode: 420
            name: tenants-fluent-bit-config-ss
          name: tenants-fluent-bit-config-ss
      fluentBconfigmapOcs:
        - configMap:
            defaultMode: 420
            name: tenants-fluent-bit-config-ocs
          name: tenants-fluent-bit-config-ocs
      fluentBconfigmapUrs:
        - configMap:
            defaultMode: 420
            name: tenants-fluent-bit-config-urs
          name: tenants-fluent-bit-config-urs           
    mounts:
      log:
        - name: log
          mountPath: /opt/genesys/logs/volume
        - name: log
          mountPath: /logs
      jsonLog:
        - name: fluent-logs
          mountPath: "/opt/genesys/logs/JSON"         
      stdoutUrsLog:
        - name: fluenturs-logs
          mountPath: "/opt/genesys/logs/URS"
      stdoutSsLog:
        - name: fluentss-logs
          mountPath: "/opt/genesys/logs/SS"
      stdoutOcsLog:
        - name: fluentocs-logs
          mountPath: "/opt/genesys/logs/OCS"
      stdoutCsLog:
        - name: fluentcs-logs
          mountPath: "/opt/genesys/logs/confserv"
 
      fbJsonLog:
        - name: fluent-logs
          mountPath: "/mnt/logs"         
      fbstdoutUrsLog:
        - name: fluenturs-logs
          mountPath: "/mnt/logs"
      fbstdoutSsLog:
        - name: fluentss-logs
          mountPath: "/mnt/logs"
      fbstdoutOcsLog:
        - name: fluentocs-logs
          mountPath: "/mnt/logs"
      fbstdoutCsLog:
        - name: fluentcs-logs
          mountPath: "/mnt/logs"         
 
      fluentBconfigmap:
          - mountPath: /fluent-bit/etc/
            name: tenants-fluent-bit-config
      fluentBconfigmapCs:
          - mountPath: /fluent-bit/etc/
            name: tenants-fluent-bit-config-cs     
      fluentBconfigmapSs:
          - mountPath: /fluent-bit/etc/
            name: tenants-fluent-bit-config-ss   
      fluentBconfigmapOcs:
          - mountPath: /fluent-bit/etc/
            name: tenants-fluent-bit-config-ocs     
      fluentBconfigmapUrs:
          - mountPath: /fluent-bit/etc/
            name: tenants-fluent-bit-config-urs

You can deploy Persistent Volume/Persistent Volume Claim (PV/PVC) in two ways:

  1. Enable ReadWriteOnce (RWO) from the tenant Helm chart, which maintains unique PVCs for each pod/replica from the same Tenant.
  2. Enable ReadWriteMany (RWX) from the tenant-monitor Helm chart, which has multiple Tenant pods sharing the same PVC.

To enable RWO PV/PVC logging from individual Tenant pods in Statefulset, make the following modifications to override the values in the tenant-values.yaml file. RWO Persistent Volume is disabled by default.

logging
 ....
  volumes: 
    logPvc:
        enabled: "true"
        logClaimSize: "5Gi"
        accessModes: "ReadWriteOnce"
        logStorageClass: "<storage class>"   #Replace the storage class that's relevant to the Openshift Cluster

Enabling GWS integration (the secret should be created as described in the Genesys Authentication backend secrets section):

tenant:
…
gws:
    # enable: Enable GWS registeration about tenant
    # tls: Enable/Disable Secure connection to GWS
    # authEndpoint: GWS auth end point
    # envEndpoint: GWS env end point for Registeration
    # db: Pass DB infromation for GWS to connect to PSQL DB for read and store data
    enable: "true"
    tls: false
    authEndpoint: "gauth-auth.gauth.svc.cluster.local"
    envEndpoint: ""
    db:
      enable: "false"
      read: "false"
      init: "false"
             secrets:
……….
gws:
      enabled: true 
      user:
        secretName: "/opt/genesys/gauth-client-id/clientid"
      pwd:
        secretName: "/opt/genesys/gauth-client-token/clientsecret"
      volumes:
        - name: gauth-client-id
          secret:
            secretName: gauthclientid    
        - name: gauth-client-token
          secret:
            secretName: gauthclientsecret     
      mounts:          
        - name: gauth-client-id
          readOnly: true
          mountPath: "/opt/genesys/gauth-client-id"
        - name: gauth-client-token
          readOnly: true
          mountPath: "/opt/genesys/gauth-client-token"

To mount the PVC to store Tenant logs, make the following modifications to override the values in the tenant-values.yaml file:

.....
 
tenant:
  ...
    logging:
    mounts:
      log:
        - name: log
          mountPath: /opt/genesys/logs/volume
        - name: log
          mountPath: /logs

Validate the deployment

Content coming soon