MirrorCache

A reverse proxy supporting multiple cache policies, customized rules, configuration hot reloading. For mirror sites and also personal users!

docs | demo server

Features

• Cache your dependencies in limited space with LRU/TTL cache policies
• Reload updated config without restarting the program
• Support customized rules
• Expose metrics like cache hit rate per rule, task count and more
• Support multiple storage backend (memory, local filesystem and S3)

Quick start

1. Download the latest release
   or install with cargo:

    cargo install mirror-cache
   

   or pull from docker hub:

    docker pull seanchao/mirror-cache
   

2. Prepare a configuration file

3. Try it out! e.g

   pip install -i http://localhost:9000 requests
   
   conda install -c http://localhost:9000 requests
   conda config --set custom_channels.pytorch http://localhost:9000/anaconda/cloud/ && conda install -c pytorch -y --download-only -v torchtext