Quick Installation
Here is a set of Octopus service is deployed quickly through Helm 3.0.0+.
So you need to have the support of helm, please refer to install Helm
Then you can deploy the service by adding the chart repository.
# Helm 3.0.0+
# add repository into helm repos
$ helm repo add ocharts https://open-intelligence.github.io/charts/
# fetch octopus chart
$ helm fetch ocharts/octopus
After the download is complete, there will be octopus-x.x.x.tgz
compressed files in the current directory, and the ./octopus
directory will be obtained after decompressing the files.
Next, you need to modify the values.yaml
configuration file in the ./octopus
directory. The configuration file contains the configuration information of each subcomponent. Only the items that must be configured in combination with the current installation environment are listed here. Other configuration items can be based on actual needs Adjustment.
# values.yaml
# Interface service configuration items
rest-server:
# Database configuration
storage:
# Default root account password
rootPassword: "root"
dataVolume:
# Data persistence path
hostPath: ""
# Labels of the database running node
nodeSelector:
octopus.openi.pcl.cn/rest-server-storage: "yes"
docker:
# Mirror warehouse address
registry: ""
# Mirror warehouse username
username: ""
# Mirror warehouse password
password: ""
volumes:
- name: ghome
# No need to change
mountPath: /ghome
# Data set directory
# Each user will automatically generate a unique subdirectory for this purpose, and load the task container when the task starts
hostPath: /ghome
- name: gmodel
# No need to change
mountPath: /gmodel
# Model directory
# Each user will automatically generate a unique subdirectory for this purpose, and load the task container when the task starts
hostPath: /gmodel
# Function plug-in configuration items
rest-server-plugin:
# Database configuration
storage:
# Default root account password
rootPassword: "root"
dataVolume:
# Data persistence path
hostPath: ""
# Labels of the database running node
nodeSelector:
octopus.openi.pcl.cn/rest-server-plugin-storage: "yes"
service:
# Maximum running time of debug type tasks(sec)
debugJobMaxRunTime: 7200
sharehosts:
# rest-server.volumes.[0].hostPath
shareDirectory: "/ghome"
common:
# Turn off jupyterLab request body size limit, turn on the limit for individual requests to be limited to 1M
disableJpyLabRequestBodyLimit: "true"
# Task management service configuration items
taskset-core:
# Database configuration
storage:
# Default root account password
rootPassword: "root"
dataVolume:
# Data persistence path
hostPath: ""
# Labels of the database running node
nodeSelector:
octopus.openi.pcl.cn/taskset-core-storage: "yes"
# Log service configuration items
log-factory:
es:
replicaCount: 1
# ElasticSearch Index sharding configuration
index:
number_of_shards: 5
number_of_replicas: 2
volumes:
# Data persistence path
esDataPath: ""
# Backup persistent path
esBackupPath: ""
ES_JAVA_OPTS: ""
resources: {
}
# Labels of the database running node
nodeSelector:
octopus.openi.pcl.cn/log-service-es: "yes"
filebeat:
# Resource limit
resources: {
limits: {
cpu: "1000m",
memory: "2Gi"
},
requests: {
cpu: "1000m",
memory: "2Gi"
}
}
# Monitoring service configuration items
prometheus:
prometheus:
nodeSelector:
octopus.openi.pcl.cn/prometheus: "yes"
retentionDuration: 365d
volumes:
# Data persistence path
dataPath: ""
After configuration, perform the installation:
# install service
$ helm install octopus ./octopus --values ./octopus/values.yaml