StressTest
Description
Receives online stress testing requests from the O&M platform and forwards the corresponding operation to the specified node of a training job. It must be called only after a training job has successfully completed its execution and iteration, to ensure that the job has been registered with ClusterD. This API represents a manual O&M operation. Before calling this API, ensure that the server environment is normal.
Deliver the online stress testing command after the training iteration is normal.
Prototype
rpc StressTest(StressTestParam) returns (Status) {}
Input Parameters
Parameter |
Type (Defined by Protobuf) |
Description |
|---|---|---|
StressTest |
message StressTestParam { string jobID = 1; map<string, StressOpList> stressParam = 2; repeated int64 allNodesOps = 3; } message StressOpList { repeated int64 ops = 1; } |
StressTestParam.jobID: job ID StressTestParam.stressParam: node and operation receiving the user-issued stress testing instruction. key indicates the node name, and value indicates the stress testing operation to be performed on the node. StressTestParam.allNodesOps: stress testing operation to be performed on all nodes. The priority of allNodesOps is higher than that of stressParam. 0 indicates AIC stress testing, and 1 indicates P2P stress testing. StressOpList.ops: stress testing operation to be performed on the node. 0 indicates AIC stress testing, and 1 indicates P2P stress testing. |
Return Value
Parameter |
Type (Defined by Protobuf) |
Description |
|---|---|---|
Status |
message Status{ int32 code = 1; string info = 2; } |
Status.code: return code
Status.info: return information |