What Are the Recommended Specifications for CPTS Resource Groups?
In Cloud Performance Test Service (CPTS), resource group nodes used for pressure tests must be created in Cloud Container Engine (CCE) in advance. The specifications of the nodes are the same as those of CCE.
When creating a CCE resource group, configure the node specifications as follows:
- Minimum specifications: 4 vCPUs and 8 GB memory.
- Maximum specifications: 32 vCPUs and 128 GB memory.
For a CPTS test project, if the number of concurrent users is less than 1000, you can use a shared resource group for testing and do not have to create a resource group. For a JMeter test project, you need to create a resource group regardless of the number of concurrent users.
Usage Recommendations
- Do not run any applications or perform other functions on nodes used for resource groups in pressure tests. Otherwise, applications may run abnormally.
- Create nodes with specifications based on the number of concurrent users for pressure tests. The nodes must be created in CCE. Only the CCE clusters of v1.15 and earlier versions are supported. You are advised to select Tunnel network as the cluster network model. When you create nodes, select EulerOS as the OS.
At least two nodes must be created. One node is used as the management node and the other is used as the execution node. In a pressure test resource group, an execution node executes pressure tests and provides performance data, while the management node manages the execution node. If a pressure test needs to be performed on an external service, bind an elastic IP address to the execution node.
Table 1 Recommended specifications Number of Concurrent Users
Specifications
Quantity
0-5000
Management node: 4 vCPUs | 8 GB
1
Execution node: 4 vCPUs | 8 GB
1
5001-10,000
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
1
10,001-20,000
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
2
20,001-30,000
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
3
30,001-40,000
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
4
40,001-50,000
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
5
More than 50,001
Management node: 4 vCPUs | 8 GB
1
Execution node: 8 vCPUs | 16 GB
n
NOTE:Each execution node (8 vCPUs | 16 GB) supports 10,000 concurrent users.
NoteThe preceding specifications are for reference only. In actual pressure tests, resource specifications are affected by factors such as think time, protocol type, the size and number of requests and responses, response time, and result verification. Adjust the specifications based on your requirements.
- Usage Recommendations