sdadams

6 ms·

1.
by sdadams 4y ago
No error correcting codes currently. Just checksums and redundant copies. Currently the supported checksum algorithms are crc32, md5, and sha1.
2.
by sdadams 4y ago
There is a FUSE implementation of ZFS for Linux but it is still in development. Porting it at this stage could be tricky.
3.
by sdadams 4y ago
Thanks! I'll try to remember this for any future blog entries.
4.
by sdadams 4y ago
The OpenBSD FUSE implementation is in base so it should already be well audited. I don't plan to add "concatenation" as you have described, however this can, in theory, be approximated by layering muxfs on top of multiple RAI
5.
by sdadams 4y ago
Porting ZFS to OpenBSD has been attempted before. See this blog post from 2013: https://flak.tedunangst.com/post/ZFS-on-OpenBSD
6.
by sdadams 4y ago
The muxfs source code is a lot smaller than that of ZFS so if security is a concern to you then you might find muxfs easier to audit than ZFS. It also compiles quickly so could be a good match for a source-based system. This said I never
7.
by sdadams 4y ago
In another project of mine I considered using extended attributes to tag files into categories. With extended attributes not being universally supported and easily overlooked the conclusion I came to was that I should store that informatio
8.
by sdadams 4y ago
OpenBSD removed support for extended attributes.
9.
by sdadams 4y ago
OpenBSD's FUSE does not implement multithreading.
10.
by sdadams 4y ago
(1) I believe that stable inode numbers are possible with FUSE but they must be implemented by the FUSE driver. Mirroring over a network is not the intended use case. The mirrors exist to provide data redundancy for use by the muxfs drive
11.
by sdadams 4y ago
Author here. A big thank you to you all for your interest in muxfs! I will try to answer all of your questions as best I can.