[Bug][v1.3 release test] ChatQnA failed with vLLM on rocm

[chensuyue]

Priority

P1-Stopper

OS type

Ubuntu

Hardware type

GPU-AMD

Installation method

• Pull docker images from hub.docker.com
• Build docker images from source
• Other
• N/A

Deploy method

• Docker
• Docker Compose
• Kubernetes Helm Charts
• Kubernetes GMC
• Other
• N/A

Running nodes

Single Node

What's the version?

main branch
a222d1c

Description

Both ChatQnA/tests/test_compose_faqgen_vllm_on_rocm.sh and ChatQnA/tests/test_compose_vllm_on_rocm.sh failed on tei-rerank microservice test.
https://github.com/opea-project/GenAIExamples/actions/runs/14420869112/job/40549840866
https://github.com/opea-project/GenAIExamples/actions/runs/14460523041/job/40552209543?pr=1811

Reproduce steps

cd GenAIExamples/ChatQnA/tests
bash test_compose_faqgen_vllm_on_rocm.sh
bash test_compose_vllm_on_rocm.sh

Raw log

[ tei-rerank ] HTTP status is not 200. Received status was 000
Error: Process completed with exit code 1.

Attachments

No response

[chensuyue]

@chyundunovDatamonsters

[chensuyue]

Found a new issue during test, https://github.com/opea-project/GenAIExamples/actions/runs/14483954532/job/40626052953?pr=1799#step:6:936

[chensuyue]

Test pass after change ROCm test machine.