Quick Deployment

This section helps users efficiently deploy the "LiteLLM, Unified AI Management Gateway" solution. When deploying this solution with one click, follow the steps and instructions in this section to complete the deployment quickly.

1. OClick "Start Deployment" to jump to the RFS deployment page.
   Figure 1 Selecting a template
   
2. Click "Next" to confirm basic configurations and set ECS password.
   Figure 2 Configuring parameters
   

   Table1 Single-Cloud Server Deployment Parameter description

   Parameter	Description	Default Value
   solution_name	Solution name: 4-24 characters, supports lowercase letters, numbers, - (hyphen), must start with a lowercase letter.	headroom-claude-code
   ecs_flavor	ECS instance type: x1.8u.16g (8vCPUs 16GiB) or higher recommended.	x1.8u.16g
   ecs_password	ECS password: 8-26 characters, must contain at least three of the following: uppercase letters, lowercase letters, numbers, and special characters.	/
   db_password	PostgreSQL password: 8-26 characters. Used by LiteLLM for internal database (virtual keys, spending tracking).	/
   master_key	LiteLLM master key: Must start with 'sk-'. Used for Admin UI login and API authentication.	/
   salt_key	API Key encryption key: Cannot be changed after setting, can be auto-generated.	Optional
   system_disk_size	System disk size (GB): Ultra-high IO type, value range: 40-1024. Default: 200.	200
   bandwidth_size	Elastic public network bandwidth (Mbit/s): Billed by traffic, value range: 1-300. Default: 200.	200
   charging_mode	Billing mode: postPaid (pay-as-you-go) or prePaid (subscription). Default: postPaid.	postPaid
   charging_unit	Subscription period type: month or year, only valid for prePaid mode.	month(Not effective)
   charging_period	Subscription period: 1-9 (months) or 1-3 (years), only valid for prePaid mode.	1(Not effective)

   Table2 HA Deployment in CCE Container Parameter description

   Parameter	Description	Default Value
   resource_name_prefix	Resource name prefix. Naming rule: {resource_name_prefix}-{resource_english_name}. Range: 4-24 characters, supports lowercase letters, digits, and hyphens (-), must start with a lowercase letter.	ha-litellm
   bandwidth_size	Elastic public IP bandwidth size, billed by traffic. Unit: Mbit/s, range: 1-300.	300
   litellm_version	LiteLLM Proxy image version. The template will automatically assemble it as docker.litellm.ai/berriai/litellm:{version}.	1.87.0
   litellm_master_key	LiteLLM master key. Must start with sk-. This is the Bearer Token used to call the proxy externally.	/
   litellm_salt_key	LiteLLM Salt Key, used to encrypt sensitive data in the database. Recommended to use a random string of 8+ characters. Max 24 characters.	/
   cce_cluster_flavor	CCE Turbo cluster flavor. Default is cce.s2.small (small-scale multi-master CCE cluster, max 50 nodes).	cce.s2.small
   cce_node_pool_password	CCE cluster node password, used for cluster node login.	/
   cce_node_pool_flavor	CCE cluster node ECS instance flavor. Recommended to meet LiteLLM's official production recommendation of 4C8G per instance.	c7n.2xlarge.2
   cce_node_pool_count	Initial number of nodes in the CCE node pool.	/
   rds_flavor	RDS for PostgreSQL primary/standby instance flavor. Default 8U16G (dedicated).	rds.pg.x1.2xlarge.2.ha
   pgsql_admin_password	PostgreSQL administrator password.	/
   redis_capacity	Distributed Cache Service Redis instance capacity. Options: 1GB-64GB, default 8GB.	8
   redis_password	Redis password.	/
   charging_mode	Billing mode. Options: postPaid (pay-per-use), prePaid (monthly/yearly subscription).	postPaid
   charging_unit	Subscription period unit. Only effective when charging_mode=prePaid.	month (not effective)
   charging_period	Subscription period. Only effective when charging_mode=prePaid.	1 (not effective)

3. On the Configure Stack page, select rf_admin_trust from the Agency drop-down list and click Next. This step is optional if you use an account (HUAWEI ID) or use an IAM user in the admin user group.
   Figure 3 Configuring a stack
   
4. On the Confirm Configurations page, confirm the configurations and click Create Execution Plan.
   Figure 4 Confirming the configurations
   
5. In the displayed Create Execution Plan dialog box, enter an execution plan name and click OK.
   Figure 5 Creating an execution plan
   
6. Wait until the status of the execution plan changes to Available and then click Deploy in the Operation column. In the displayed dialog box, click Execute.
   Figure 6 Execution plan page
   
   Figure 7 Confirming the execution plan
   
7. (Optional) If you select the yearly/monthly billing mode and your account balance is insufficient, log in to the Billing Center to manually pay for the order. You can refer to Table2 Resource planning costs (yearly/monthly) to see the total price.
8. Wait until the message "Apply required resource success" is displayed on the Events tab page. This means the deployment is complete. The deployment takes about 10 minutes, which will be delayed by network fluctuations.
   Figure 8 Resources created
   

   9.Click "Output" to view relevant prompts in the "Value" section, and wait 10 minutes before starting to use the service.

   Figure 9 View "Value"