Cluster Configuration (Helm Chart Variables)

Root

The fully qualified host name, via which you will use to access your self-hosted instance.

The organization-level Public Key. This key MUST have permissions to create projects within your account.

The name of the MongoDB Database.

The username of the account which will be created, and then subsequently used for auth against the MongoDB cluster.

The name of the PostgreSQL database.

The username of the account which will be created, and then subsequently used for auth against the PSQL instance.

Cluster

The name of your UserWise cluster. This can be anything (default: “userwise-hosted-cluster”)

The version of Kubernetes to use, primarily during initial construction of your infrastructure.

Cluster Frontend Pod

Defines the number of web server processes to run per pod. Generally speaking, a minimum of 4 processes is recommended, to help alleviate any over-saturation of CPU cores.

cluster.frontend_pod.requests.pod should generally be equal to the number of processes (1:1).

Defines the number of threads to allow per process. Generally speaking a value of 5-32 is ideal.

Pod PSQL Connections: num_processes * num_threads_per_process

Enables the fine-tuning of the number of CPU cores to assign per front end pod.

Enables the fine-tuning of the amount of memory to assign per front end pod.

Enables the fine-tuning of the amount of storage to assign per front end pod.

Pods generally log to STDOUT, ephemeral storage usage is kept to a minimum.

Enables the fine-tuning of the number of CPU cores to assign per front end pod.

Enables the fine-tuning of the amount of memory to assign per front end pod.

Enables the fine-tuning of the amount of storage to assign per front end pod.

Pods generally log to STDOUT, ephemeral storage usage is kept to a minimum.

Cluster Worker Pod

Defines the number of default_worker queue processes to start per worker pod.

This queue is used for any uncategorized worker jobs.

Defines the number of player_content_send_worker queue processes to start per worker pod.

This queue is used for any delivery of background content (e.g. emails & push notifications)

Defines the number of player_data_worker queue processes to start per worker pod.

This queue is used for any player data processing (e.g. new sessions, event assignment, attribute assignment, etc.)

Enables the fine-tuning of the number of CPU cores to assign per worker pod.

Enables the fine-tuning of the amount of memory to assign per worker pod.

Enables the fine-tuning of the amount of storage to assign per worker pod.

Pods generally log to STDOUT, ephemeral storage usage is kept to a minimum.

Enables the fine-tuning of the number of CPU cores to assign per worker pod.

Enables the fine-tuning of the amount of memory to assign per worker pod.

Enables the fine-tuning of the amount of storage to assign per worker pod.

Pods generally log to STDOUT, ephemeral storage usage is kept to a minimum.

Cluster Horizontal Pod Autoscaler (HPA)

Allows enabling the HPA for both the frontend and worker pods.

NOTE: If you enable this, it is recommended to disable the cluster.static_pod_scaling.enabled configuration.

Defines the minimum number of replicas for the frontend pods.

Defines the maximum number of replicas for the frontend pods.

Defines the percent usage of available CPU cores (across all frontend pods); used to trigger a scale event.

Defines the minimum number of replicas for the worker pods.

Defines the maximum number of replicas for the worker pods.

Defines the percent usage of available CPU cores (across all worker pods); used to trigger a scale event.

Cluster Static Pod Scaling

Allows enabling static pod scaling functionality for both the frontend and worker pods.

NOTE: If you enable this, it is recommended to disable the cluster.horizontal_pod_autoscaler.enabled configuration.

Defines the static number of frontend pods to run within the cluster.

Defines the static number of worker pods to run within the cluster.

AWS

While we offer hosting of the actual cluster between either AWS or GCP, there are some configuration options that are required when using GCP as well (denoted by an [required]).

Allows enabling/disabling of the AWS hosting integration.

NOTE: If you enable this, it is required that you disable gcp.cluster_hosting_enabled

Your AWS account id.

This field is required even when hosting on GCP.

Your AWS preferred region.

This field is required even when hosting on GCP. Ideally, if hosting on GCP, this should be geographically close to your selected GCP region.

AWS TLS Certificate which is related to the host defined at the root of this file.

AWS ECR

The name of the AWS ECR repository, which will be used to store your UserWise built images.

AWS EKS

Defines the minimum default nodegroup size.

Defines the maximum default nodegroup size.

Defines the desired size of the default node group.

AWS IAM

The name of the AWS IAM user account to create.

This service account is used primarily to access SES, SNS, & S3.

AWS MemoryDB

It is currently recommended to setup Elasticache in-place of memory db. This can be updated within the self hosted repo, or manually.

An update is in the plans to bring this as the default.

The name of the redis instance.

The version of redis to use.

The number of shards. Should always be a single node cluster.

The number of replicas per shard.

AWS RDS

The name/identifier of the RDS instance.

Allows enabling/disabling of RDS multi-az deployment (recommended)

The RDS instance class to use for this RDS instance.

Defines the minimal storage available for the PSQL instance.

AWS S3

The S3 bucket name to create, which is used for media hosting, as well as exports hosting/access.

AWS VPC

The name used when creating the AWS VPC.

The IP Address CIDR Block to use for the VPC

The IP ranges used when defining the VPC private subnets.

The IP ranges used when defining the VPC public subnets.

GCP

Allows enabling/disabling of the GCP hosting integration.

NOTE: If you enable this, it is required that you disable aws.cluster_hosting_enabled

Defines the GCP project id to setup all infrastructure under.

The unique identifier of the service account created during infrastructure setup.

The display name of the service account created during infrastructure setup.

The GCP region to host the infrastructure under.

GCP Artifact Registry

The name of the container image repository to setup within Artifact Registry.

This repository is used to store self-hosted custom built UserWise container images.

GCP CSQL

The CSQL database instance tier to use.

The name of the CSQL database instance.

Defines the number of days to store the transaction logs for.

Defines the number of backups to retain at once.

GCP GKE

Configures the various components to enable for monitoring within GCP GKE.

Configures the various components to enable for logging within GCP GKE.

The CIDR range which should be used by the cluster infrastructure.

The CIDR range which should be used by the cluster’s services.

GCP MemoryStore

The name of the memory store instance.

The GCP MemoryStore tier to use for this instance.

Defines the minimum size of the memorystore store (recommended minimum of 2GB).

Defines the redis version to use within the memorystore instance.

Defines the MemoryStore instance’s persistence mode.

Defins the period interval which is used for backups.

GCP VPC

Defines the GCP VPC network’s name.

Atlas

Defines the MongoDB cluster’s name within Atlas.

Defines the provider to use within atlas.

NOTE: This MUST be either AWS or GCP, dependent on your hosting environment. This is used to help create secure connections between your account and the Atlas account (VPC Peering).

Defines the region used within the provider’s services.

NOTE: This MUST be equivalent to the region you are hosting your services on.

Defines the minimum instance size within your MongoDB cluster.

Defines the maximum instance size within your MongoDB cluster.

Defines the CIDR block used by the cluster.

Minimal Configuration:

{
  "host": "uw.mydomain.com",

  "atlas_public_key": "",
  "mongo_db_name": "userwise",
  "mongo_username": "userwisemongouser",
  "psql_db_name": "userwise",
  "psql_username": "userwisepsqluser",

  "cluster": {
    "name": "userwise-hosted-cluster",
    "k8s_version": "1.28",

    "frontend_pod": {
      "processes": {
        "num_processes": 8,
        "num_threads_per_process": 10
      },
      "requests": {
        "cpu": 8,
        "memory": "8Gi",
        "ephemeral_storage": "2Gi"
      },
      "limits": {
        "cpu": 8,
        "memory": "8Gi",
        "ephemeral_storage": "2Gi"
      }
    },

    "worker_pod": {
      "queue_configuration": {
        "default_worker": 1,
        "player_content_send_worker": 1,
        "player_data_worker": 3
      },
      "requests": {
        "cpu": 5,
        "memory": "5Gi",
        "ephemeral_storage": "2Gi"
      },
      "limits": {
        "cpu": 5,
        "memory": "5Gi",
        "ephemeral_storage": "2Gi"
      }
    },

    "horizontal_pod_autoscaler": {
      "enabled": true,
      "min_frontend_replicas": 1,
      "max_frontend_replicas": 10,
      "frontend_cpu_threshold_percentage": 50,
      "min_worker_replicas": 1,
      "max_worker_replicas": 5,
      "worker_cpu_threshold_percentage": 50
    },

    "static_pod_scaling": {
      "enabled": false,
      "frontend_replicas": 1,
      "worker_replicas": 1
    }
  },

  "aws": {
    "cluster_hosting_enabled": false,
    "account_id": "784192393659",
    "region": "us-west-1",
    "certificate_arn": "",

    "ecr": {
      "repository_name": "userwise-hosted-artifacts"
    },

    "eks": {
      "min_size": 1,
      "max_size": 3,
      "desired_size": 1
    },

    "iam": {
      "service_account_name": "userwise-service-account"
    },

    "memory_db": {
      "cluster_name": "userwise-hosted-redis",
      "engine_version": "6.2",
      "num_shards": 1,
      "num_replicas_per_shard": 0
    },

    "rds": {
      "identifier": "userwise-hosted-psql",
      "multi_az": true,
      "instance_class": "db.t3.medium",
      "ensure_allocated_storage": 30
    },

    "s3": {
      "bucket_name": "uw-hosted-bucket"
    },

    "vpc": {
      "name": "userwise-vpc",
      "cidr": "172.31.0.0/16",
      "private_subnets": ["172.31.10.0/24", "172.31.20.0/24", "172.31.30.0/24", "172.31.40.0/24"],
      "public_subnets": ["172.31.100.0/24", "172.31.110.0/24", "172.31.120.0/24", "172.31.130.0/24"]
    }
  },

  "gcp": {
    "cluster_hosting_enabled": true,
    "project_id": "uw-project",
    "service_account_id": "userwise-service",
    "service_account_display_name": "UserWise Infra Service Account",
    "region": "us-central1",

    "artifact_registry": {
      "repository_name": "userwise-hosted-artifacts"
    },

    "csql": {
      "tier": "db-g1-small",
      "instance_name": "userwise-hosted-psql",
      "transaction_log_retention_days": 5,
      "num_retained_backups": 5
    },

    "gke": {
      "monitoring_config": ["SYSTEM_COMPONENTS", "APISERVER", "CONTROLLER_MANAGER", "SCHEDULER"],
      "logging_config": ["SYSTEM_COMPONENTS", "APISERVER", "CONTROLLER_MANAGER", "SCHEDULER", "WORKLOADS"],
      "cluster_cidr": "10.90.0.0/16",
      "services_cidr": "10.91.0.0/16"
    },

    "memorystore": {
      "instance_name": "userwise-hosted-redis",
      "tier": "STANDARD_HA",
      "memory_size_gb": 1,
      "redis_version": "REDIS_6_X",
      "persistence_mode": "RDB",
      "rdb_snapshot_period": "SIX_HOURS"
    },

    "vpc": {
      "network_name": "userwise-vpc"
    }
  },

  "atlas": {
    "cluster_name": "userwise-hosted-cluster",
    "provider": "GCP",
    "provider_region": "CENTRAL_US",
    "min_instance_size": "M10",
    "max_instance_size": "M60",
    "cidr": "172.20.0.0/18"
  }
}

Cluster Configuration (Helm Chart Variables)

Root

Cluster

Cluster Frontend Pod

Cluster Worker Pod

Cluster Horizontal Pod Autoscaler (HPA)

Cluster Static Pod Scaling

AWS

AWS ECR

AWS EKS

AWS IAM

AWS MemoryDB

AWS RDS

AWS S3

AWS VPC

GCP

GCP Artifact Registry

GCP CSQL

GCP GKE

GCP MemoryStore

GCP VPC

Atlas

Minimal Configuration:

Related content