# High-Availability Configuration

    When deploying Erda via Helm Chart package, you can choose deployment mode by setting the global parameter global.size:

    • Set global.size as demo, then all components of Erda will be deployed in the minimalist way (low-resource configuration and single-instance copy), which is suitable for trial environment.
    • Set global.size as prod, then the core components of Erda will be deployed in a high-availability way (high-resource configuration and multi-instance copy), which is suitable for production environment.

    This article will introduce the configuration for production deployment (high-availability mode).

    # Configurable Parameter for High-Availability Deployment

    The values.yaml (opens new window) file in Erda Helm Chart defines a large number of configuration parameters, which can be modified according to the actual situation for high-availability deployment.

    Parameter Description Default Value
    Global
    global.size The deployment mode (demo and prod are supported), and the high-availability mode is set as prod -
    global.image.repository Set the image repository address and users who cannot access the external network need to modify the configuration to a private repository on the internal network and upload the image required for Erda deployment to the private repository before deployment "registry.erda.cloud/erda"
    global.image.imagePullPolicy Set image pull policy "IfNotPresent"
    global.image.imagePullSecrets Setting not required if image is not pulled from private repository, otherwise it needs to be set as secrets to access the private image repository -
    global.domain The domain name bound to the current Erda cluster "erda.io"
    Cassandra
    cassandra.capacity Set the single-node storage capacity of Cassandra, which is scalable according to the actual cluster and business volume "1000Gi"
    cassandra.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    cassandra.resources.requests.cpu Set the CPU request value of the Cassandra instance pod "2"
    cassandra.resources.requests.memory Set the memory request value of the Cassandra instance pod "4Gi"
    cassandra.resources.limits.cpu Set the CPU limit value of the Cassandra instance pod "4"
    cassandra.resources.limits.memory Set the memory limit value of the Cassandra instance pod "16Gi"
    cassandra.racks Rack name list (names cannot be repeated), and the number of racks in the list equal to the number of Cassandra nodes - name: rack1
    - name: rack2
    - name: rack3
    Elasticsearch
    elasticsearch.replicas Set the number of Elasticsearch cluster nodes 3
    elasticsearch.capacity Set the single-node storage capacity of Elasticsearch, which is scalable according to the actual cluster and business volume "1000Gi"
    elasticsearch.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    elasticsearch.numberOfMasters When deploying an Elasticsearch cluster in high-availability mode, the number of Elasticsearch instances is generally set over half of the cluster node number to avoid split-brain in the Elasticsearch cluster 2
    elasticsearch.javaOpts Set the environment variable JAVA_OPTS of Elasticsearch (Java heap size is recommended to be set as 0.75 * resources.limits.memory) "-Xms6144m -Xmx6144m"
    elasticsearch.resources.requests.cpu Set the CPU request value of the Elasticsearch instance pod "2"
    elasticsearch.resources.requests.memory Set the memory request value of the Elasticsearch instance pod "4Gi"
    elasticsearch.resources.limits.cpu Set the CPU limit value of the Elasticsearch instance pod "4"
    elasticsearch.resources.limits.memory Set the memory limit value of the Elasticsearch instance pod "8Gi"
    etcd
    etcd.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    etcd.capacity Set the single-node storage capacity of etcd, which is scalable according to the actual cluster and business volume "32Gi"
    etcd.resources.requests.cpu Set the CPU request value of the etcd instance pod "1"
    etcd.resources.requests.memory Set the memory request value of the etcd instance pod "2Gi"
    etcd.resources.limits.cpu Set the CPU limit value of the etcd instance pod "4"
    etcd.resources.limits.memory Set the memory limit value of the etcd instance pod "8Gi"
    Zookeeper
    zookeeper.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    zookeeper.capacity Set the single-node storage capacity of Zookeeper, which is scalable according to the actual cluster and business volume "32Gi"
    zookeeper.resources.requests.cpu Set the CPU request value of the Zookeeper instance pod "100m"
    zookeeper.resources.requests.memory Set the memory request value of the Zookeeper instance pod "256Mi"
    zookeeper.resources.limits.cpu Set the CPU limit value of the Zookeeper instance pod "1"
    zookeeper.resources.limits.memory Set the memory limit value of the Zookeeper instance pod "512Mi"
    Kafka
    kafka.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    kafka.capacity Set the single-node storage capacity of Kafka, which is scalable according to the actual cluster and business volume "32Gi"
    kafka.javaOpts Set the environment variable JAVA_OPTS of Kafka (Java heap size is recommended to be set as 0.75 * resources.limits.memory) "-Xms6144m -Xmx6144m"
    kafka.resources.requests.cpu Set the CPU request value of the Kafka instance pod "2"
    kafka.resources.requests.memory Set the memory request value of the Kafka instance pod "4Gi"
    kafka.resources.limits.cpu Set the CPU limit value of the Kafka instance pod "4"
    kafka.resources.limits.memory Set the memory limit value of the Kafka instance pod "8Gi"
    KMS
    kms.replicas Set the number of KMS instance replicas 2
    kms.resources.requests.cpu Set the CPU request value of the KMS instance pod "500m"
    kms.resources.requests.memory Set the memory request value of the KMS instance pod "1Gi"
    kms.resources.limits.cpu Set the CPU limit value of the KMS instance pod "1"
    kms.resources.limits.memory Set the memory limit value of the KMS instance pod "2Gi"
    Redis
    redis.redisFailover.redis.replicas Set the number of Redis replicas for active/passive switch among Redis instances 2
    redis.redisFailover.redis.resources.requests.cpu Set the CPU request value of the Redis pod "150m"
    redis.redisFailover.redis.resources.requests.memory Set the memory request value of the Redis pod "1Gi"
    redis.redisFailover.redis.resources.limits.cpu Set the CPU limit value of the Redis pod "300m"
    redis.redisFailover.redis.resources.limits.memory Set the memory limit value of the Redis pod "2Gi"
    redis.redisFailover.sentinel.replicas Set the number of Redis Sentinel replicas 3
    Registry
    registry.storageClassName Set the Kubernetes StorageClass object corresponding to the storage volume "dice-local-volume"
    registry.capacity Set the single-node storage capacity of Registry, which is scalable according to the actual cluster and business volume "1000Gi"
    registry.resources.requests.cpu Set the CPU request value of the Registry instance pod "500m"
    registry.resources.requests.memory Set the memory request value of the Registry instance pod "512Mi"
    registry.resources.limits.cpu Set the CPU limit value of the Registry instance pod "1"
    registry.resources.limits.memory Set the memory limit value of the Registry instance pod "1Gi"
    registry.networkMode If the value is "host", set the registry container network as host mode -
    registry.custom.nodeName Registry shares the same name with the node deployed in host mode -
    registry.custom.nodeIP Registry shares the same IP address of the node deployed in host mode -
    Sonar
    sonar.resources.requests.cpu Set the CPU request value of the Sonar instance pod "750m"
    sonar.resources.requests.memory Set the memory request value of the Sonar instance pod "1536Mi"
    sonar.resources.limits.cpu Set the CPU limit value of the Sonar instance pod "1500m"
    sonar.resources.limits.memory Set the memory limit value of the Sonar instance pod "3Gi"
    volume-provisioner
    volume-provisioner.provisioner.local.path Local Volume takes this mount point as the storage volume source /data
    volume-provisioner.provisioner.nfs.path FS Volume takes this mount point as the storage volume source /netdata
    volume-provisioner.resources.requests.cpu Set the CPU request value of the volume-provisioner instance pod "10m"
    volume-provisioner.resources.requests.memory Set the memory request value of the volume-provisioner instance pod "10Mi"
    volume-provisioner.resources.limits.cpu Set the CPU limit value of the volume-provisioner instance pod "100m"
    volume-provisioner.resources.limits.memory Set the memory limit value of the volume-provisioner instance pod "256Mi"
    Erda
    erda.clusterName The identifier of the Kubernetes cluster where Erda is deployed "local-cluster"
    erda.operator.resources.requests.cpu Set the CPU request value of the erda-operator instance pod "10m"
    erda.operator.resources.requests.memory Set the memory request value of the erda-operator instance pod "10Mi"
    erda.operator.resources.limits.cpu Set the CPU limit value of the erda-operator instance pod "100m"
    erda.operator.resources.limits.memory Set the memory limit value of the erda-operator instance pod "128Mi"
    erda.clusterConfig.protocol Declare the request protocol of the current Erda cluster, HTTP/HTTPS http
    erda.clusterConfig.clusterType Erda cluster identifier, such as Kubernetes and EDAS kubernetes
    erda.component.admin.replicas The number of erda admin component replicas 2
    erda.component.admin.resources.cpu The CPU request value of the erda admin component instance pod "100m"
    erda.component.admin.resources.mem The memory request value of the erda admin component instance pod "128Mi"
    erda.component.admin.resources.max_cpu The CPU limit value of the erda admin component instance pod "200m"
    erda.component.admin.resources.max_mem The memory limit value of erda admin component instance pod "256Mi"
    erda.component.clusterManager.replicas The number of erda clusterManager component replicas 2
    erda.component.clusterManager.resources.cpu The CPU request value of the erda clusterManager component instance pod "100m"
    erda.component.clusterManager.resources.mem The memory request value of the erda clusterManager component instance pod "128Mi"
    erda.component.clusterManager.resources.max_cpu The CPU limit value of the erda clusterManager component instance pod "200m"
    erda.component.clusterManager.resources.max_mem The memory limit value of the erda clusterManager component instance pod "256Mi"
    erda.component.collector.replicas The number of erda collector component replicas 2
    erda.component.collector.resources.cpu The CPU request value of the erda collector component instance pod "100m"
    erda.component.collector.resources.mem The memory request value of the erda collector component instance pod "128Mi"
    erda.component.collector.resources.max_cpu The CPU limit value of the erda collector component instance pod "1"
    erda.component.collector.resources.max_mem The memory limit value of the erda collector component instance pod "1024Mi"
    erda.component.coreServices.replicas The number of erda coreServices component replicas 2
    erda.component.coreServices.resources.cpu The CPU request value of the erda coreServices component instance pod "100m"
    erda.component.coreServices.resources.mem The memory request value of the erda coreServices component instance pod "128Mi"
    erda.component.coreServices.resources.max_cpu The CPU limit value of the erda coreServices component instance pod "300m"
    erda.component.coreServices.resources.max_mem The memory limit value of the erda coreServices component instance pod "512Mi"
    erda.component.hepa.replicas The number of erda hepa component replicas 2
    erda.component.hepa.resources.cpu The CPU request value of the erda hepa component instance pod "100m"
    erda.component.hepa.resources.mem The memory request value of the erda hepa component instance pod "512Mi"
    erda.component.hepa.resources.max_cpu The CPU limit value of the erda hepa component instance pod "500m"
    erda.component.hepa.resources.max_mem The memory limit value of the erda hepa component instance pod -
    erda.component.monitor.replicas The number of erda monitor component replicas 2
    erda.component.monitor.resources.cpu The CPU request value of the erda monitor component instance pod "100m"
    erda.component.monitor.resources.mem The memory request value of the erda monitor component instance pod "128Mi"
    erda.component.monitor.resources.max_cpu The CPU limit value of the erda monitor component instance pod "1"
    erda.component.monitor.resources.max_mem The memory limit value of the erda monitor component instance pod "512Mi"
    erda.component.msp.replicas The number of erda msp component replicas 2
    erda.component.msp.resources.cpu The CPU request value of the erda msp component instance pod "100m"
    erda.component.msp.resources.mem The memory request value of the erda msp component instance pod "128Mi"
    erda.component.msp.resources.max_cpu The CPU limit value of the erda msp component instance pod "1"
    erda.component.msp.resources.max_mem The memory limit value of the erda msp component instance pod "512Mi"
    erda.component.openapi.replicas The number of erda openapi component replicas 2
    erda.component.openapi.resources.cpu The CPU request value of the erda openapi component instance pod "100m"
    erda.component.openapi.resources.mem The memory request value of the erda openapi component instance pod "128Mi"
    erda.component.openapi.resources.max_cpu The CPU limit value of the erda openapi component instance pod "500m"
    erda.component.openapi.resources.max_mem The memory limit value of the erda openapi component instance pod "512Mi"
    erda.component.scheduler.replicas The number of erda scheduler component replicas 2
    erda.component.scheduler.resources.cpu The CPU request value of the erda scheduler component instance pod "100m"
    erda.component.scheduler.resources.mem The memory request value of the erda scheduler component instance pod "128Mi"
    erda.component.scheduler.resources.max_cpu The CPU limit value of the erda scheduler component instance pod "1"
    erda.component.scheduler.resources.max_mem The memory limit value of the erda scheduler component instance pod "2048Mi"
    erda.component.streaming.replicas The number of erda streaming component replicas 2
    erda.component.streaming.resources.cpu The CPU request value of the erda streaming component instance pod "100m"
    erda.component.streaming.resources.mem The memory request value of the erda streaming component instance pod "128Mi"
    erda.component.streaming.resources.max_cpu The CPU limit value of the erda streaming component instance pod "1500m"
    erda.component.streaming.resources.max_mem The memory limit value of the erda streaming component instance pod "1024Mi"
    erda.component.ui.replicas The number of erda ui component replicas 2
    erda.component.ui.resources.cpu The CPU request value of the erda ui component instance pod "200m"
    erda.component.ui.resources.mem The memory request value of the erda ui component instance pod "256Mi"
    erda.component.ui.resources.max_cpu The CPU limit value of the erda ui component instance pod "1"
    erda.component.ui.resources.max_mem The memory limit value of the erda ui component instance pod "512Mi"
    erda.component.ucAdaptor.replicas The number of erda ucAdaptor component replicas 2
    erda.component.ucAdaptor.resources.cpu The CPU request value of the erda ucAdaptor component instance pod "100m"
    erda.component.ucAdaptor.resources.mem The memory request value of the erda ucAdaptor component instance pod "64Mi"
    erda.component.ucAdaptor.resources.max_cpu The CPU limit value of the erda ucAdaptor component instance pod "200m"
    erda.component.ucAdaptor.resources.max_mem The memory limit value of the erda ucAdaptor component instance pod -
    erda.component.uc.replicas The number of erda uc component replicas 2
    erda.component.uc.resources.cpu The CPU request value of the erda uc component instance pod "10m"
    erda.component.uc.resources.mem The memory request value of the erda uc component instance pod "100Mi"
    erda.component.uc.resources.max_cpu The CPU limit value of the erda uc component instance pod "1"
    erda.component.uc.resources.max_mem The memory limit value of the erda uc component instance pod "2048Mi"
    erda.component.cmp.replicas The number of erda cmp component replicas 2
    erda.component.cmp.resources.cpu The CPU request value of the erda cmp component instance pod "100m"
    erda.component.cmp.resources.mem The memory request value of the erda cmp component instance pod "128Mi"
    erda.component.cmp.resources.max_cpu The CPU limit value of the erda cmp component instance pod "200m"
    erda.component.cmp.resources.max_mem The memory limit value of the erda cmp component instance pod -
    erda.component.analyzerAlert.resources.cpu The CPU request value of the erda analyzerAlert component instance pod "100m"
    erda.component.analyzerAlert.resources.mem The memory request value of the erda analyzerAlert component instance pod "128Mi"
    erda.component.analyzerAlert.resources.max_cpu The CPU limit value of the erda analyzerAlert component instance pod "1"
    erda.component.analyzerAlert.resources.max_mem The memory limit value of the erda analyzerAlert component instance pod "1024Mi"
    erda.component.analyzerAlertTask.resources.cpu The CPU request value of the erda analyzerAlertTask component instance pod "100m"
    erda.component.analyzerAlertTask.resources.mem The memory request value of the erda analyzerAlertTask component instance pod "128Mi"
    erda.component.analyzerAlertTask.resources.max_cpu The CPU limit value of the erda analyzerAlertTask component instance pod "1"
    erda.component.analyzerAlertTask.resources.max_mem The memory limit value of the erda analyzerAlertTask component instance pod "2048Mi"
    erda.component.analyzerErrorInsight.resources.cpu The CPU request value of the erda analyzerErrorInsight component instance pod "100m"
    erda.component.analyzerErrorInsight.resources.mem The memory request value of the erda analyzerErrorInsight component instance pod "128Mi"
    erda.component.analyzerErrorInsight.resources.max_cpu The CPU limit value of the erda analyzerErrorInsight component instance pod "1"
    erda.component.analyzerErrorInsight.resources.max_mem The memory limit value of the erda analyzerErrorInsight component instance pod "2048Mi"
    erda.component.analyzerErrorInsightTask.resources.cpu The CPU request value of the erda analyzerErrorInsightTask component instance pod "100m"
    erda.component.analyzerErrorInsightTask.resources.mem The memory request value of the erda analyzerErrorInsightTask component instance pod "128Mi"
    erda.component.analyzerErrorInsightTask.resources.max_cpu The CPU limit value of the erda analyzerErrorInsightTask component instance pod "1"
    erda.component.analyzerErrorInsightTask.resources.max_mem The memory limit value of the erda analyzerErrorInsightTask component instance pod "2048Mi"
    erda.component.analyzerMetrics.resources.cpu The CPU request value of the erda analyzerMetrics component instance pod "100m"
    erda.component.analyzerMetrics.resources.mem The memory request value of the erda analyzerMetrics component instance pod "128Mi"
    erda.component.analyzerMetrics.resources.max_cpu The CPU limit value of the erda analyzerMetrics component instance pod "1"
    erda.component.analyzerMetrics.resources.max_mem The memory limit value of the erda analyzerMetrics component instance pod "2048Mi"
    erda.component.analyzerMetricsTask.resources.cpu The CPU request value of the erda analyzerMetricsTask component instance pod "100m"
    erda.component.analyzerMetricsTask.resources.mem The memory request value of the erda analyzerMetricsTask component instance pod "128Mi"
    erda.component.analyzerMetricsTask.resources.max_cpu The CPU limit value of the erda analyzerMetricsTask component instance pod "1"
    erda.component.analyzerMetricsTask.resources.max_mem The memory limit value of the erda analyzerMetricsTask component instance pod "2048Mi"
    erda.component.analyzerTracing.resources.cpu The CPU request value of the erda analyzerTracing component instance pod "100m"
    erda.component.analyzerTracing.resources.mem The memory request value of the erda analyzerTracing component instance pod "128Mi"
    erda.component.analyzerTracing.resources.max_cpu The CPU limit value of the erda analyzerTracing component instance pod "500m"
    erda.component.analyzerTracing.resources.max_mem The memory limit value of the erda analyzerTracing component instance pod "1024Mi"
    erda.component.analyzerTracingTask.resources.cpu The CPU request value of the erda analyzerTracingTask component instance pod "100m"
    erda.component.analyzerTracingTask.resources.mem The memory request value of the erda analyzerTracingTask component instance pod "128Mi"
    erda.component.analyzerTracingTask.resources.max_cpu The CPU limit value of the erda analyzerTracingTask component instance pod "1"
    erda.component.analyzerTracingTask.resources.max_mem The memory limit value of the erda analyzerTracingTask component instance pod "2048Mi"
    erda.component.actionRunnerScheduler.resources.cpu The CPU request value of the erda actionRunnerScheduler component instance pod "100m"
    erda.component.actionRunnerScheduler.resources.mem The memory request value of the erda actionRunnerScheduler component instance pod "128Mi"
    erda.component.actionRunnerScheduler.resources.max_cpu The CPU limit value of the erda actionRunnerScheduler component instance pod "300m"
    erda.component.actionRunnerScheduler.resources.max_mem The memory limit value of the erda actionRunnerScheduler component instance pod -
    erda.component.clusterAgent.resources.cpu The CPU request value of the erda clusterAgent component instance pod "100m"
    erda.component.clusterAgent.resources.mem The memory request value of the erda clusterAgent component instance pod "128Mi"
    erda.component.clusterAgent.resources.max_cpu The CPU limit value of the erda clusterAgent component instance pod "1"
    erda.component.clusterAgent.resources.max_mem The memory limit value of the erda clusterAgent component instance pod "1024Mi"
    erda.component.clusterDialer.resources.cpu The CPU request value of the erda clusterDialer component instance pod "100m"
    erda.component.clusterDialer.resources.mem The memory request value of the erda clusterDialer component instance pod "128Mi"
    erda.component.clusterDialer.resources.max_cpu The CPU limit value of the erda clusterDialer component instance pod "2"
    erda.component.clusterDialer.resources.max_mem The memory limit value of the erda clusterDialer component instance pod "2048Mi"
    erda.component.dicehub.resources.cpu The CPU request value of the erda dicehub component instance pod "100m"
    erda.component.dicehub.resources.mem The memory request value of the erda dicehub component instance pod "128Mi"
    erda.component.dicehub.resources.max_cpu The CPU limit value of the erda dicehub component instance pod "150m"
    erda.component.dicehub.resources.max_mem The memory limit value of the erda dicehub component instance pod "1024Mi"
    erda.component.dop.resources.cpu The CPU request value of the erda dop component instance pod "100m"
    erda.component.dop.resources.mem The memory request value of the erda dop component instance pod "128Mi"
    erda.component.dop.resources.max_cpu The CPU limit value of the erda dop component instance pod "1"
    erda.component.dop.resources.max_mem The memory limit value of the erda dop component instance pod "2048Mi"
    erda.component.eventbox.resources.cpu The CPU request value of the erda eventbox component instance pod "100m"
    erda.component.eventbox.resources.mem The memory request value of the erda eventbox component instance pod "128Mi"
    erda.component.eventbox.resources.max_cpu The CPU limit value of the erda eventbox component instance pod "2"
    erda.component.eventbox.resources.max_mem The memory limit value of the erda eventbox component instance pod "2560Mi"
    erda.component.filebeat.resources.cpu The CPU request value of the erda filebeat component instance pod "100m"
    erda.component.filebeat.resources.mem The memory request value of the erda filebeat component instance pod "128Mi"
    erda.component.filebeat.resources.max_cpu The CPU limit value of the erda filebeat component instance pod "1"
    erda.component.filebeat.resources.max_mem The memory limit value of the erda filebeat component instance pod "512Mi"
    erda.component.gittar.resources.cpu The CPU request value of the erda gittar component instance pod "100m"
    erda.component.gittar.resources.mem The memory request value of the erda gittar component instance pod "128Mi"
    erda.component.gittar.resources.max_cpu The CPU limit value of the erda gittar component instance pod "1"
    erda.component.gittar.resources.max_mem The memory limit value of the erda gittar component instance pod "1536Mi"
    erda.component.pipeline.replicas The number of erda pipeline component replicas 2
    erda.component.pipeline.resources.cpu The CPU request value of the erda pipeline component instance pod "100m"
    erda.component.pipeline.resources.mem The memory request value of the erda pipeline component instance pod "128Mi"
    erda.component.pipeline.resources.max_cpu The CPU limit value of the erda pipeline component instance pod "1"
    erda.component.pipeline.resources.max_mem The memory limit value of the erda pipeline component instance pod "1536Mi"
    erda.component.telegraf.resources.cpu The CPU request value of the erda telegraf component instance pod "100m"
    erda.component.telegraf.resources.mem The memory request value of the erda telegraf component instance pod "128Mi"
    erda.component.telegraf.resources.max_cpu The CPU limit value of the erda telegraf component instance pod "500m"
    erda.component.telegraf.resources.max_mem The memory limit value of the erda telegraf component instance pod "512Mi"
    erda.component.telegrafApp.resources.cpu The CPU request value of the erda telegrafApp component instance pod "100m"
    erda.component.telegrafApp.resources.mem The memory request value of the erda telegrafApp component instance pod "128Mi"
    erda.component.telegrafApp.resources.max_cpu The CPU limit value of the erda telegrafApp component instance pod "500m"
    erda.component.telegrafApp.resources.max_mem The memory limit value of the erda telegrafApp component instance pod "512Mi"
    erda.component.telegrafPlatform.resources.cpu The CPU request value of the erda telegrafPlatform component instance pod "100m"
    erda.component.telegrafPlatform.resources.mem The memory request value of the erda telegrafPlatform component instance pod "128Mi"
    erda.component.telegrafPlatform.resources.max_cpu The CPU limit value of the erda telegrafPlatform component instance pod "1"
    erda.component.telegrafPlatform.resources.max_mem The memory limit value of the erda telegrafPlatform component instance pod "1536Mi"
    erda.component.orchestrator.resources.cpu The CPU request value of the erda orchestrator component instance pod "100m"
    erda.component.orchestrator.resources.mem The memory request value of the erda orchestrator component instance pod "128Mi"
    erda.component.orchestrator.resources.max_cpu The CPU limit value of the erda orchestrator component instance pod "1"
    erda.component.orchestrator.resources.max_mem The memory limit value of the erda orchestrator component instance pod "256Mi"

    # Configuration Parameter for Core Data Storage Component

    For clusters with different node number, please refer to the following parameter configuration of Cassandra, Elasticsearch, Kafka and other storage components in high-availability mode.

    Cluster Size 0~50 Nodes 50~100 Nodes 100~200 Nodes 200~300 Nodes 300+ Nodes
    Cassandra
    cassandra.resources.requests.cpu "1" "2" "4" "4" "4"
    cassandra.resources.requests.memory "6Gi" "12Gi" "16Gi" "16Gi" "16Gi"
    cassandra.resources.limits.cpu "2" "4" "6" "6" "6"
    cassandra.resources.limits.memory "12Gi" "16Gi" "24Gi" "24Gi" "24Gi"
    cassandra.capacity 512G 1T 1.5T 1.5T 2T
    cassandra.racks - name: rack1
    - name: rack2
    - name: rack3
    - name: rack1
    - name: rack2
    - name: rack3
    - name: rack1
    - name: rack2
    - name: rack3
    - name: rack1
    - name: rack2
    - name: rack3
    - name: rack4
    - name: rack5
    - name: rack1
    - name: rack2
    - name: rack3
    - name: rack4
    - name: rack5
    - name: rack6
    - name: rack7
    Elasticsearch
    elasticsearch.resources.requests.cpu "1" "2" "4" "4" "4"
    elasticsearch.resources.requests.memory "4Gi" "8Gi" "16Gi" "16Gi" "16Gi"
    elasticsearch.resources.limits.cpu "2" "4" "6" "6" "6"
    elasticsearch.resources.limits.memory "8Gi" "16Gi" "24Gi" "24Gi" "24Gi"
    elasticsearch.capacity 512G 768G 1T 1.5T 1.5T
    elasticsearch.replicas 3 3 3 3 5
    elasticsearch.numberOfMasters 2 2 2 2 3
    Kafka
    kafka.resources.requests.cpu "0.5" "1" "1" "1" "2"
    kafka.resources.requests.memory "1Gi" "2Gi" "2Gi" “2Gi" "4Gi"
    kafka.resources.limits.cpu "1" "2" "2" "2" "4"
    kafka.resources.limits.memory "2Gi" "4Gi" "4Gi" "4Gi" "8Gi"
    kafka.capacity 150G 150G 200G 300G 300G

    # How to Connect to Existing Middleware

    Erda relies on a variety of middleware, such as Elasticsearch, MySQL, Kafka and Registry, some of which can be configured as existing instance without installation.

    Tips

    • Currently only external MySQL is supported for connection, and connection for other middleware (such as Kafka and Elasticsearch) is coming soon.
    • When connecting to external MySQL, the version needs to be at least 5.7.9, and the default character should be utf8m64 for database.

    You can modify the values.yaml in Erda Chart package by adding the following fields to connect to external MySQL:

    mysql:
      enabled: false
      custom:
        address:      #  eg: 192.168.100.100
        port:         #  eg: 3306
        database:     #  eg: erda
        user:         #  eg: root
        password:     #  eg: HasdDwqwe23#
    
    

    After adding the above configuration, the MySQL database provided by user can be directly applied with no need to deploy MySQL component.

    The specific parameters are as follows:

    Parameter Description
    mysql.enabled Set as false when connecting to external MySQL
    mysql.custom.address Connect to the MySQL host address provided by the user
    mysql.custom.port Connect to the MySQL host port provided by the user
    mysql.custom.databases Connect to the MySQL database provided by the user
    mysql.custom.user Connect to the user name of MySQL database provided by the user
    mysql.custom.password Connect to the password corresponding to the user name of MySQL database provided by the user

    # How to Save Private Configuration

    Using the parameter configuration in the values.yaml file to deploy Helm Chart is the easiest way, but it cannot meet all situations. You can adjust parameter configuration in the following ways:

    • Option 1 (recommended): Add modified parameters to the customized values.yaml file and run -f to specify the file when install or upgrade Helm.
    • Option 2: Run --set to set parameter value when install or upgrade Helm. However, --set cannot make persistence configuration, which may lead to inconsistent parameter settings of upgrade and installation.
    • Option 3: Modify parameter values of the values.yaml file in the Helm Chart package. But in the case of a huge amount of parameters, it is difficult to quickly confirm whether the parameters need to be, or have been modified.