[CICD] Optimize kv cache image size
🚀 Feature Description and Motivation
The image size is super large now, we need to reduce the size a little bit.
FROM ubuntu:22.04
RUN apt-get update && \
DEBIAN_FRONTEND=noninteractive apt-get install -y \
python3 python3-pip \
iproute2 net-tools \
ibverbs-utils libibverbs-dev \
libnl-3-dev libnl-route-3-dev \
rdmacm-utils libibverbs1 rdma-core \
wget \
&& rm -rf /var/lib/apt/lists/*
#RUN pip3 install --no-cache-dir InfiniStore
#
RUN wget https://test-files.pythonhosted.org/packages/f5/57/7013d0deee8b5a0e8cfd5a51bcc8be2084fc0ab8589586fb5e19687efe99/infinistore-0.2.41-cp310-cp310-manylinux_2_28_x86_64.whl
RUN pip3 install infinistore-0.2.41-cp310-cp310-manylinux_2_28_x86_64.whl
CMD ["infinistore"]
Use Case
Reduce the container image size
Proposed Solution
No response
Seems the 2nd build for gid patch result in larger image.. But we only install wget and download the whl. whl is just 9MiB
hi @Jeffwan I want to solve this problem, but the describedin the issue is about InfiniStore. Is the Dockerfile for KVCache https://github.com/vllm-project/aibrix/blob/main/build/container/Dockerfile.kvcache?
or accturlly want to resolve InfiniStore image size?
Could you perhaps give me some guidance? thanks!
@yyzxw sorry for late response. Dockerfile.kvcache is not the right dockerfile. that's the image to sync kv cache information to redis.
Here, we focus more on the infinistore image itself. I think we didn't check in the Dockerfile earlier. Could you create a new one instead? like Dockerfile.infinistore. Please test the latest version the and version have issues.