Implementation:Triton inference server Server L0 Memory Growth Test

Knowledge Sources	Triton Inference Server
Domains	Testing, Memory_Management
Last Updated	2026-02-13 17:00 GMT

Overview

QA test script for validating server memory growth over extended inference runs in the Triton Inference Server.

Description

This test validates that Triton Inference Server does not exhibit unbounded memory growth during sustained inference workloads. It launches the server, sends a large number of inference requests over an extended period, and monitors resident memory consumption to detect memory leaks. The test fails if memory usage grows beyond an acceptable threshold relative to the baseline measurement taken at startup.

Usage

Run as part of the Triton QA test suite. Requires a GPU-enabled Docker environment with pre-generated test models.

Code Reference

Source Location

• Repository: Triton Inference Server
• File: qa/L0_memory_growth/test.sh
• Lines: 1-314

Signature

#!/bin/bash
source ../common/util.sh
# Test orchestration for server memory growth over extended inference runs

Import

source ../common/util.sh

I/O Contract

Inputs

Name	Type	Required	Description
DATADIR	env var	No	Test data directory
MODEL_REPO	directory	Yes	Test model repository

Outputs

Name	Type	Description
exit code	int	0 on success, 1 on failure
test logs	files	Server and test output logs

Usage Examples

Running the Test

cd qa/L0_memory_growth/
bash test.sh

Related Pages

• Environment:Triton_inference_server_Server_GPU_CUDA_Runtime