1fuse-overlayfs(1) General Commands Manual fuse-overlayfs(1)
2
3
4
6 fuse-overlayfs - overlayfs FUSE implementation
7
8
9
11 mounting
12 fuse-overlayfs [-f] [--debug] [-o OPTS] MOUNT_TARGET
13
14
15 unmounting
16 fusermount -u mountpoint
17
18
19
21 fuse-overlayfs provides an overlayfs FUSE implementation so that it can
22 be used since Linux 4.18 by unprivileged users in an user namespace.
23
24
25
27 --debug Enable debugging mode, can be very noisy.
28
29
30 -o lowerdir=low1[:low2...] A list of directories separated by :.
31 Their content is merged.
32
33
34 -o upperdir=upperdir A directory merged on top of all the lowerdirs
35 where all the changes done to the file system will be written.
36
37
38 -o workdir=workdir A directory used internally by fuse-overlays, must
39 be on the same file system as the upper dir.
40
41
42 -o uidmapping=UID:MAPPED-UID:LEN[,UID2:MAPPED-UID2:LEN2] -o gidmap‐
43 ping=GID:MAPPED-GID:LEN[,GID2:MAPPED-GID2:LEN2] Specifies the dynamic
44 UID/GID mapping used by fuse-overlayfs when reading/writing files to
45 the system.
46
47
48 The fuse-overlayfs dynamic mapping is an alternative and cheaper way to
49 chown'ing the files on the host to accommodate the user namespace set‐
50 tings.
51
52
53 It is useful to share the same storage among different user namespaces
54 and counter effect the mapping done by the user namespace itself, and
55 without requiring to chown the files.
56
57
58 For example, given on the host two files like:
59
60
61 $ stat -c %u:%g lower/a lower/b 0:0 1:1
62
63
64 When we run in a user namespace with the following configuration: $ cat
65 /proc/self/uid_map
66 0 1000 1
67 1 110000 65536
68
69
70 We would see:
71
72
73 $ stat -c %u:%g merged/a merged/b 65534:65534 65534:65534
74
75
76 65534 is the overflow id used when the UID/GID is not known inside the
77 user namespace. This happens because both users 0:0 and 1:1 are not
78 mapped.
79
80
81 In the above example, if we mount the fuse-overlayfs file system using:
82 -ouidmapping=0:1000:1:1:110000:65536,gidmap‐
83 ping=0:1000:1:1:110000:65536, which is the namespace configuration
84 specified on a single line, we'd see from the same user namespace:
85
86
87 $ stat -c %u:%g merged/a merged/b 0:0 1:1
88
89
90 Those are the same IDs visible from outside the user namespace.
91
92
93 -o squash_to_root Every file and directory is owned by the root user
94 (0:0).
95
96
97 -o squash_to_uid=uid -o squash_to_gid=gid Every file and directory is
98 owned by the specified uid or gid.
99
100
101 It has higher precedence over squash_to_root.
102
103
104 -o static_nlink Set st_nlink to the static value 1 for all directories.
105
106
107 This can be useful for higher latency file systems such as NFS, where
108 counting the number of hard links for a directory with many files can
109 be a slow operation. With this option enabled, the number of hard links
110 reported when running stat for any directory is 1.
111
112
113
115 fuse(8), mount(8), user_namespaces(7)
116
117
118
120 The fuse-overlayfs command is available from https://github.com/con‐
121 tainers/fuse-overlayfs under GNU GENERAL PUBLIC LICENSE Version 3 or
122 later.
123
124
125
126 User Commands fuse-overlayfs(1)