Docker构建运行时遇TypeError：opentelemetry proto文件重复

错误信息 #

TypeError: Couldn't build proto file into descriptor pool: duplicate file name opentelemetry/proto/common/v1/common.proto

requirements.txt 内容 #

numpy~=1.26.4
oauthlib==3.2.2
openai==1.12.0
pandas==1.5.3
pandas-stubs==1.2.0.39
requests==2.32.0
requests-oauthlib==1.3.0
uvicorn==0.23.0
anyio==3.5.0
tenacity==8.2.3
plotly==5.13.0
scipy
scikit-learn
pydantic~=1.10.12
tiktoken==0.5.2
regex
PyPDF2==3.0.1
docx2txt==0.8
python-docx==0.8.11
python-multipart==0.0.9
pybase64==1.2.3
slowapi==0.1.8
nltk==3.8.1
Jinja2==3.1.4
pytailwindcss==0.1.4
pdf2image==1.16.3
pypdf==3.17.0
langchain==0.1.14
langchain-community==0.0.30
langchain-openai==0.0.4
typing_extensions==4.8.0
unstructured==0.6.11
python-pptx==0.6.23
APScheduler==3.10.4
python-socketio==5.8.0
websockets==11.0.3
google-cloud-aiplatform==1.44.0
pandasai==2.1
bs4==0.0.1
matplotlib==3.7.2
matplotlib-inline==0.1.6
PyMySQL==1.1.1
SQLAlchemy==1.4.49
pypandoc-binary==1.12
qdrant-client==1.9.1
chromadb==0.5.0
protobuf==4.25.3

报错调用栈 #

from pandasai.ee.vectorstores.pinecone import Pinecone  
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/pandasai/ee/vectorstores/__init__.py", line 5, in <module>
June 07, 2024 at 13:08 (UTC+5:30)   from .chroma import ChromaDB
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/pandasai/ee/vectorstores/chroma.py", line 5, in <module>   
June 07, 2024 at 13:08 (UTC+5:30)   import chromadb 
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/chromadb/__init__.py", line 5, in <module> 
June 07, 2024 at 13:08 (UTC+5:30)   from chromadb.auth.token_authn import TokenTransportHeader  usw2-dev-taskgpt-core-cn
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/chromadb/auth/token_authn/__init__.py", line 20, in <module>
June 07, 2024 at 13:08 (UTC+5:30)   from chromadb.telemetry.opentelemetry import (
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/chromadb/telemetry/opentelemetry/__init__.py", line 12, in <module>
June 07, 2024 at 13:08 (UTC+5:30)   from opentelemetry.exporter.otlp.proto.grpc.trace_exporter import OTLPSpanExporter  
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/opentelemetry/exporter/otlp/proto/grpc/trace_exporter/__init__.py", line 24, in <module>   
June 07, 2024 at 13:08 (UTC+5:30)   from opentelemetry.exporter.otlp.proto.common.trace_encoder import (
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/opentelemetry/exporter/otlp/proto/common/trace_encoder.py", line 16, in <module>
June 07, 2024 at 13:08 (UTC+5:30)   from opentelemetry.exporter.otlp.proto.common._internal.trace_encoder import (
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/opentelemetry/exporter/otlp/proto/common/_internal/__init__.py", line 31, in <module>
June 07, 2024 at 13:08 (UTC+5:30)   from opentelemetry.proto.common.v1.common_pb2 import (  
June 07, 2024 at 13:08 (UTC+5:30)   File "/venv/lib/python3.12/site-packages/opentelemetry/proto/common/v1/common_pb2.py", line 17, in <module>

解决方案 #

• 排查依赖冲突：问题源于多个依赖（如chromadb、pandasai）重复引入了OpenTelemetry的proto文件，或protobuf版本与依赖不兼容。
• 固定兼容版本：尝试将protobuf降级到4.24.x版本，同时显式指定opentelemetry-proto版本，在requirements.txt中添加：

  opentelemetry-proto==1.24.0
  protobuf==4.24.4
  

• 清理Docker缓存：执行docker builder prune清理旧构建缓存，避免残留的冲突依赖干扰。
• 移除重复包：requirements中同时存在PyPDF2和pypdf，两者功能重叠会引发依赖冲突，建议保留其中一个（优先保留pypdf==3.17.0）。
• 强制依赖统一：在Docker构建阶段使用pip install --force-reinstall -r requirements.txt，强制重新安装所有依赖确保版本一致。

内容的提问来源于stack exchange，提问作者Abhi Kanimilli