Docs

Volcengine Kubernetes Engine

Deepseek practices

Quickly deploying DeepSeek-R1 (full version) with SGLang (PD disaggregation)

This topic mainly describes how to quickly deploy a model inference service with the DeepSeek-R1 full version (PD disaggregation) based on SGLang by using Volcengine Kubernetes Engine (VKE).

Solution introduction

The solution integrates multiple acceleration technologies of Volcano Engine to quickly achieve prefill-decode (PD) disaggregation deployment for the DeepSeek-R1 large language model (LLM) in a Kubernetes cluster.

DeepSeek-R1 model: Released by the Chinese AI startup DeepSeek, the DeepSeek-R1 model is an inference model trained based on DeepSeek-V3-Base. The model is good at mathematical, code, and natural language inference, which has gained wide attention due to its open source availability, algorithm innovation, and cost-effectiveness.

PD disaggregation technology: The technology divides the inference process into two phases, which are deployed in different hardware resources to optimize performance:

Prefill: A compute-intensive phase.

Decode: A low-compute phase.

Core advantages of the solution

Improved resource utilization: Separately handling Prefill and Decode phases. They respectively require high and low computing power.

Increased throughput: New requests are being processed in the Prefill phase while the old ones are being handled in the Decode phase.

Reduce latency: The wait time of concurrent requests is reduced, improving overall efficiency.

Deployment information

Model: DeepSeek-R1-671B

Inference engine: SGLang

Instance type: ecs.hpcpni3ln.45xlarge

Prefill: ecs.hpcpni3ln.45xlarge instances * 2

Decode: ecs.hpcpni3ln.45xlarge instances * 4

Supported capabilities

Parallel strategy: PD disaggregation, Data Parallelism (DP), Tensor Parallelism (TP), and Expert Parallelism (EP).

Data precision: FP8

Observability: Volcengine Managed Service for Prometheus (VMP) and Application Performance Management Plus (APMPlus).

Access: API Gateway (APIG)

Performance evaluation

The following table compares the performance of the open-source setup with that of the optimized template.

tip

The delivered performance is related to machine specifications, distribution configuration, and network bandwidth. The actual test result shall prevail.

Compare category	Open source	Optimized
Inference engine	SGLang (open source)	SGLang (Volcano Engine optimized version)
Input/Output	3500/1500	3500/1500
TTFT (ms)	<=5000	<=5000
TPOT (ms)	<=50	<=50
Parallel strategy	TP	PD disaggregation, DP, TP, and EP.
Single-GPU throughput (TPS)	167.4	567

Usage notes

In this example, many add-ons are used, which are dependent on each other and have complex configurations. To achieve the expected results, you must perform operations according to the guidance in this article.

To achieve the expected results and conform to the limits of VKE, you need to perform operations by following the solution described in this article or on the recommended resources. If you need to use a different solution, contact your Volcano Engine account manager for assistance.

Do not use mGPU on GPU resources.

A full deployment depends on some beta capabilities of the Volcano Engine. Contact the Account Manager of the Volcano Engine to apply for the beta trial of the following features:

Beta feature/resource	Product involved	Description
hpcpni3ln HPC-optimized instance	GPU Elastic Compute Service	The HPC instance, providing computing power. The DeepSeek model has a large number of parameters and therefore requires a GPU instance with large memory, e.g., ecs.hpcpni3ln.45xlarge.
Capability for mounting local disks to nodes	Volcengine Kubernetes Engine	You can configure the local disk mounting in the VKE cluster.

Prerequisites

Creating an HPC cluster

The HPC cluster can achieve the logical isolation of high-performance computing GPU instances (computing nodes). Instances in the same cluster are interconnected by Remote Direct Memory Access (RDMA) networks. For details, refer to Creating High-Performance Compute Clusters

tip

The region and availability zone where the high-performance compute cluster is located need to be consistent with those of the VKE nodes.

Creating a VKE cluster

The following parameters need to be configured as described in the table when creating a cluster by using VKE. For details on other parameters, see Creating clusters. You can also use the managed cluster template for heterogeneous computing to quickly create a VKE cluster for this scenario.

Configuration item	Description
Cluster configuration
Kubernetes version	Select version v1.28 or later.
Network model	Select VPC-CNI.
Node pool configuration
Managed node pool	Enable managed node pool.
Availability zone	Same availability zone as the High-Performance Compute Cluster.
Instance type	Select ecs.hpcpni3ln, the GPU Accelerated specification. Ensure you already obtained permission for trial of the beta version.
High Performance Computing Cluster	Select the HPC cluster you created.
Number of nodes	6
Data volume	Configure the following two data volumes: ESSD: set capacity to 1024 GiB and the performance level to the default PL0. Local SSD: Specify /data01 as the mount directory. Warning The data01 directory has subsequent dependencies. Therefore, the directory cannot be modified or specified. Ensure the data that needs long-term storage has been backed up properly in case of computer crashes or disk damage which causes data loss.
Node labels	When deploying DeepSeek model through VKE, you need to set the node label to configure the RDMA usage mode. The label key is set to: vke.node.rdma.mode, and the label value is set to: shared.
Kubelet configs	The non-uniform memory access (NUMA) affinity policy for RDMA devices needs to be configured by using custom parameters with the Kubelet when deploying DeepSeek model in the VKE. Select the parameter topology-manager-policy and set its value to best-effort.
Addon configuration
nvidia-device-plugin	(Required) The NVIDIA device driver. It allows use of management add-ons for GPUs in containers.
rdma-device-plugin	(Required) The add-on manages RDMA network interface cards (NICs) and enables pod containers to use RDMA devices.
apig-controller	(Required) API Gateway controller provides the AI Gateway capability of cloud-native, which can complete the intelligent load balancing of the inference traffic according to different policies.
prometheus-agent	(Required) Volcengine Managed Service for Prometheus (VMP) monitoring add-on collects the metrics data of the Kubernetes cluster to VMP.
apmplus-opentelemetry-collector	(Required) Collect OpenTelemetry data and Prometheus metrics, and send them to APMPlus.
apmplus-server-agent	(Required) Send the collection service and host monitoring data in parallel to APMPlus.
cr-credential-controller	(Recommended) It can pull Container Registry (CR) images without passwords, which simplifies deployment operations.
keda	(Recommended) The event-driven autoscaler based on Kubernetes completes the autoscaler tasks of the inference service based on complex metrics.
vci-virtual-kubelet	(Recommended) It can deploy CPU-consuming workloads such as system add-ons by using Volcengine Container Instance (VCI), which eliminates the need for operation and maintenance (O&M) and does not occupy the computing resources of GPU nodes.

Procedure

Step 1: Deploy LeaderWorkerSet

In this article, LeaderWorkerSet (LWS) is used to arrange the inference loads on multiple nodes (For example, DeepSeek R1 needs to be deployed across two nodes). LWS allows you to specify a group of pods as objects to be replicated.

In the left navigation bar, select Application center > Application template. Scroll to the bottom of the page, click More application templates.

On the Application template page, find leader-worker-set. Move your cursor over this application template, and then click Deploy.

On the Deploy application page, configure the service parameters of leader-worker-set.

Configuration item	Description
Orchestration template
Application template	Select the latest Chart version of leader-worker-set.
Name	Customize the application name according to the naming requirements prompted by the interface.
Deployment environment
Cluster type	Keep the default value as Managed version.
Cluster	Select the VKE cluster created in Prerequisites.
Namespace	Select a namespace under the target cluster.
Protocol	Read and understand the agreement before selecting the agreement.

Click Confirm to start deployment.

After deploying the application, it will be displayed on the Helm page. Select the target project, cluster, and namespace, and click the name of the deployed leader-worker-set application to go to the details page. On the details page, view the basic information about the application and the created workloads, services, and other resources.

Step 2: Deploy deepseek-r1-sglang-pd

VKE provides the DeepSeek-R1 model application (PD disaggregation) based on the SGLang framework. The application can be quickly deployed through the Application template page of the VKE console.

In the Volcengine Kubernetes Engine console.

In the left navigation bar, select Application center > Application template.

On the Application template page, find deepseek-r1-sglang-pd and click Deploy.

On the Deploy application page, specify the parameters of deepseek-r1-sglang-pd.

Configuration item	Description
Orchestration template
Application template	Select the Chart version of deepseek-r1-sglang-pd. We recommend that you select the latest version.
Parameter configuration	The configuration methods for deepseek-r1-sglang-pd. Click the Edit icon to view the configuration description of various parameters in the values.yaml file or edit the YAML file.
Name	Customize the application name according to the naming requirements prompted by the interface.
Deployment environment
Cluster	Select the VKE cluster created in Prerequisites.
Namespace	Select a namespace under the target cluster.
Protocol	Read and understand the agreement before selecting the agreement.

Click Confirm to start deployment.

The applications that are already deployed can be found on the Helm page. Select the target project, cluster, and namespace and click the name of the deployed deepseek-r1-sglang-pd application to go to the details page. On the details page, view the basic information about the application and associate the application with created workloads, services, and other resources.

tip

<application name>-api-svc indicates the name of the service externally exposed by the application. You need to associate the service in the subsequent steps, such as Ingress configuration and model validation.

View the workload status of the deepseek-r1-sglang-pd application on the StatefulSet page of the VKE cluster.

In the left navigation bar in the VKE console, click Cluster. On the page that appears, click the name of the target cluster.

In the left navigation bar, click Workload > StatefulSet. Then, enter the application name deepseek-r1-sglang-pd in the search box to find the deployment and view the status.

Step 3: Configure an Ingress

Configure routing for the model through APIG Ingress. Based on Volcano Engine API Gateway (APIG), APIG Ingress provides managed cloud-native gateway capabilities and enables highly available, scalable Ingress traffic management to meet requirements for traffic stability and high availability in cloud-native scenarios.

On the left navigation bar on the cluster management page, select Services and Ingresses > Ingress, and click Create Ingress. Use API Gateway to expose the model service externally in the cluster. The parameters mentioned below need to be configured as described. Other parameters can be configured by referring to the Creating APIG Ingress through the console.

Configuration item	Description
Basic configuration
Namespace	Select the namespace where the deepseek-r1-sglang-pd application template is deployed to associate the service exposed by the application.
Ingress configuration
Ingress type	Select Cloud-native API gateway.
APIG specification settings
Gateway specifications	You can select Automatically create, or select Use existing APIG instances. tip When you are using APIG to expose the model service in the clusters, we recommend that you configure the following parameters or features (Some parameters can only be configured in the APIG console, and cannot be configured in the VKE console): Instance specification: We recommend that you select 4c8g, with two availability zones to realize high availability by default. Bandwidth limit: This option is used to set the maximum bandwidth of the Classic Load Balancer (CLB) instance working with the APIG instance. If more bandwidth is required, you can specify a custom CLB instance on the APIG console. For details, see Adding access entrance. Circuit breaking: When some GPU instances malfunction, the common practice is to take the faulty instances offline to eliminate failures. At the same time, it is necessary to retain the faulty instances for troubleshooting later. You can isolate the faulty instances by configuring a circuit breaking rule for the APIG instance. Request timeout: Timeout is not applied to API Gateway by default. We recommend that you keep the default settings. If timeout is needed, you can configure it on the API Gateway console. For details, refer to Advanced routing policy. Observability: Monitoring alert and log feature are not turned on by default for the API Gateway instances created in the VKE console. If needed, we recommend that you enable logging and monitoring/alerting for the instance in the APIG console. Instance monitoring: This can help you to detect the unexpected anomalies in the whole system chain from the gateway entry level. For example: Unhealthy back end and back-end crashes will cause the APIG to return 503 error code and specific errors. Instance logs: You can view the request errors, request time consumption, and the allocation of requests in various Pods in the back end from the access logs.
Backend upstream	We recommend that you set the Load balancing algorithm parameter to Round Robin. It is more applicable to the scenarios of upgrading back-ends if the back-end loads are approximately balanced.
Listener configuration
Protocol	Select a protocol as needed. If HTTPS is selected, you need to configure the certificate as described in the document APIG Ingress.
Forwarding configuration
Domain name	Fill in the Ingress domain name to be exposed. The A (Address) record of this domain name needs to be configured later with the public or private IP address generated when the Ingress is successfully created.
Service	Select a service that is created when you deploy the deepseek-r1-sglang-pd application template and set the port number to 8080. The name of the service is in the format of <application name>-api-svc.

(Optional) Configure authentication for an APIG.

Create consumer: Select Consumer management in the left navigation bar and click Create consumer.

Generate an API-KEY: In the consumer list, click the name of the target consumer, to enter the consumer details page. In the API Key information area, click Generate API Key.

In the pop-up dialog box, select Automatically create.

Click Confirm to generate API Key credentials.

Enable Key Auth add-on: In the left navigation bar, click Plugin Center. On the page that appears, select Key Auth and then click Create Plugin.

Effective level: Select Routing level and configure the instance, service, and route to which the add-on belongs.

Authentication configuration: Select header and enter the text "Authorization" in the textbox on the right side to indicate that API Key will be included in the Authorization header of requests.

Consumer: Select the consumer created in Step 1.

Other options can remain as default.

Obtain Ingress address.

VIP: includes both public and private IP.

Backend service: includes the backend service name and access domain name.

Configure domain name resolution.

To test the DNS in the local environment, you can implement a domain-to-IP resolution locally.

In a production environment, configure the Ingress access domain name with your domain name provider, and ensure that the A record points to the public or private IP address of the Ingress.

Step 5: Enable observability

Basic observability provides the monitoring metrics for AI resources and inference service. Deep observability provides full-link performance monitoring service for inference application.

Basic observability

In the left navigation bar of the page for a specific cluster, select Overview under the gray title Cloud native observability, then click Enable on the page to start monitoring. For more configuration details, refer to Enable observability and VKE observability.

After turning on observability, the application metrics of SGLang will be collected into the Volcengine Managed Service for Prometheus (VMP). You can use Explore to query metrics in the VMP console. An example is as follows:

There is a built-in VMP dashboard that allows you to view the metrics collected from SGLang-based applications. To do that, you can go to the VMP console, select Monitoring dashboard in the left navigation bar and search for sglang to open dashboard as shown below:

You can also use Grafana to build a custom monitor dashboard for SGLang-based applications. For more details, refer to Deploying Grafana in a VKE cluster and integrating Grafana with a workspace.

SGLang Dashboard-1747206543364.json

The effect after import is shown in the figure:

Deep observability

Search for sglang and click the service name in the results to open the monitoring dashboard page.

View metrics.

LLM metrics

On tabs such as Service Overview, LLM performance, and Token usage, you can view the core metrics of the LLM service, including number of LLM calls, token usage, response latency, TTFT, TPOT, QPS, and request error rate.

tip

TTFT and TPOT are effective only when LLMs are called with streaming-enabled requests.

The example of a streaming call is as follows:

curl localhost:8080/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "/data/models/DeepSeek-R1",
"messages": [
{
"role": "user",
"content": "hello"
}
],
"stream": true,
"stream_options": {
"include_usage": true,
"continuous_usage_stats": true
}
}'

Runtime metrics

On the Runtime monitoring tab, you can view runtime-related metrics, including CPU time, CPU usage, Memory usage, Number of GC objects, Number of threads, and Number of context switches.

View link analysis.

On the Trace Analysis tab of the SGLang application, you can search trace data using different filters. Click the target Trace ID to go to the single-trace details page, where you can view the trace information in list view.

Model verification

Warning

In the PD architecture, the system may perform compilation during the first 10 service calls in the initial phase. The compilation process in this phase may lead to long response time. After the compilation is complete, the response time of the system will return to normal, which ensures the efficiency and stability of subsequent service calls.

Method 1: Access the model by using the service

After you deploy the deepseek-r1-sglang-pd application template, the created service whose name is in the format of <application name>-api-svc will provide an HTTP API endpoint (port 8080). The endpoint can be called as you do with OpenAI APIs.

The sample code is as follows.

tip

Replace ****** in http://****** with the endpoint of your <application name>-api-svc Service.

The parameter content is what you want to type to chat with LLMs.

In the Workload section of the Helm details page that corresponds to the deepseek-r1-sglang-pd application, click the name of the created StatefulSet to go to the StatefulSet page. In the upper-right corner of the details page, click Console to log in to the pod that corresponds to the application to call the API.

Method 2: Access via APIG

When enabling the authentication feature of APIG, you need to add an authentication token in the request.

The following example is provided for reference.

tip

Replace the example.com in http://example.com with the domain name of the back-end services obtained in step 4: Configure Ingress.

Replace eyJhbGciOiJSUzI****** with the API Key generated in Step 4: configure Ingress.

The parameter content is what you want to type to chat with LLMs.

Last updated: 2026.07.24 15:50:47