Catfs is a caching filesystem written in Rust.
[](https://travis-ci.org/kahing/catfs)
[](https://crates.io/crates/catfs)
[](https://crates.io/crates/catfs)
[](https://github.com/kahing/catfs/releases/)
[](https://twitter.com/CatfsFuse)
# Overview
Catfs allows you to have cached access to another (possibily remote)
filesystem. Caching semantic is read-ahead and write-through (see
[Current Status](#current-status)). Currently it only provides a data
cache and all metadata operations hit the source filesystem.
Catfs is ALPHA software. Don't use this if you value your data.
# Installation
* On Linux, install via
[pre-built binaries](https://github.com/kahing/catfs/releases/). You
may also need to install fuse-utils first.
* Or build from source which requires [Cargo](http://doc.crates.io/).
```ShellSession
:~/catfs$ cargo install catfs
$ # optimized binary now in $HOME/.cargo/bin/catfs
```
# Usage
Catfs requires extended attributes (xattr) to be enabled on the
filesystem where files are cached to. Typically this means you need to
have `user_xattr` mount option turned on.
```ShellSession
$ catfs <from> <to> <mountpoint>
```
Catfs will expose files in `<from>` under `<mountpoint>`, and cache
them to `<to>` as they are accessed. You can use `--free` to control
how much free space `<to>`'s filesystem has.
# Benchmark
Compare using catfs to cache sshfs vs sshfs only. Topology is
laptop - 802.11n - router - 1Gbps wired - desktop. Laptop has SSD
whereas desktop has spinning rust.
Compare running catfs with two local directories on the same
filesystem with direct access. This is not a realistic use case but
should give you an idea of the worst case slowdown.
Write is twice as slow as expected since we are writing twice the
amount.
<a name="runbenchmark"></a>
To run the benchmark, do:
```ShellSession
$ sudo docker run -e SSHFS_SERVER=user@host --rm --privileged --net=host -v $PWD/target:/root/catfs/target kahing/catfs-bench
# result is written to $PWD/target
```
The docker container will need to be able to ssh to `user@host`. Typically I arrange that by mounting the ssh socket from the host
```ShellSession
$ sudo docker run -e SSHFS_OPTS="-o ControlPath=/root/.ssh/sockets/%r@%h_%p -o ControlMaster=auto -o StrictHostKeyChecking=no -o Cipher=arcfour user@host:/tmp" -e SSHFS_SERVER=user@host --rm --privileged --net=host -v $HOME/.ssh/sockets:/root/.ssh/sockets -v $PWD/target:/root/catfs/target kahing/catfs-bench
```
# License
Copyright (C) 2017 Ka-Hing Cheung
Licensed under the Apache License, Version 2.0
# Current Status
Catfs is ALPHA software. Don't use this if you value your data.
Entire file is cached if it's open for read, even if nothing is
actually read.
Data is written-through to the source and also cached for each
write. In case of non-sequential writes, `catfs` detects `ENOTSUP`
emitted by filesystems like `goofys` and falls back to flush the
entire file on `close()`. Note that in the latter case even changing
one byte will cause the entire file to be re-written.
# References
* Catfs is designed to work with [goofys](https://github.com/kahing/goofys/)
* [FS-Cache](https://www.kernel.org/doc/Documentation/filesystems/caching/fscache.txt)
provides caching for some in kernel filesystems but doesn't support
other FUSE filesystems.
* Other similar fuse caching filesystems, no idea about their completeness:
* [CacheFS](https://github.com/cconstantine/CacheFS) - written in
Python, not to be confused with FS-Cache above which is in kernel
* [fuse-cache](https://sourceforge.net/projects/fuse-cache/)
* [mcachefs](https://github.com/Doloops/mcachefs)
* [pcachefs](https://github.com/ibizaman/pcachefs)