20 This filesystem provides a fairly simple secure AFS filesystem driver. It is
21 under development and does not yet provide the full feature set. The features
22 it does support include:
24 (*) Security (currently only AFS kaserver and KerberosIV tickets).
30 It does not yet support the following AFS features:
36 (*) pioctl() system call.
43 The filesystem should be enabled by turning on the kernel configuration
46 CONFIG_AF_RXRPC - The RxRPC protocol transport
47 CONFIG_RXKAD - The RxRPC Kerberos security handler
48 CONFIG_AFS - The AFS filesystem
50 Additionally, the following can be turned on to aid debugging:
52 CONFIG_AF_RXRPC_DEBUG - Permit AF_RXRPC debugging to be enabled
53 CONFIG_AFS_DEBUG - Permit AFS debugging to be enabled
55 They permit the debugging messages to be turned on dynamically by manipulating
56 the masks in the following files:
58 /sys/module/af_rxrpc/parameters/debug
59 /sys/module/afs/parameters/debug
66 When inserting the driver modules the root cell must be specified along with a
67 list of volume location server IP addresses:
71 insmod kafs.o rootcell=cambridge.redhat.com:172.16.18.73:172.16.18.91
73 The first module is the AF_RXRPC network protocol driver. This provides the
74 RxRPC remote operation protocol and may also be accessed from userspace. See:
76 Documentation/networking/rxrpc.txt
78 The second module is the kerberos RxRPC security driver, and the third module
79 is the actual filesystem driver for the AFS filesystem.
81 Once the module has been loaded, more modules can be added by the following
84 echo add grand.central.org 18.7.14.88:128.2.191.224 >/proc/fs/afs/cells
86 Where the parameters to the "add" command are the name of a cell and a list of
87 volume location servers within that cell, with the latter separated by colons.
89 Filesystems can be mounted anywhere by commands similar to the following:
91 mount -t afs "%cambridge.redhat.com:root.afs." /afs
92 mount -t afs "#cambridge.redhat.com:root.cell." /afs/cambridge
93 mount -t afs "#root.afs." /afs
94 mount -t afs "#root.cell." /afs/cambridge
96 Where the initial character is either a hash or a percent symbol depending on
97 whether you definitely want a R/W volume (hash) or whether you'd prefer a R/O
98 volume, but are willing to use a R/W volume instead (percent).
100 The name of the volume can be suffixes with ".backup" or ".readonly" to
101 specify connection to only volumes of those types.
103 The name of the cell is optional, and if not given during a mount, then the
104 named volume will be looked up in the cell specified during insmod.
106 Additional cells can be added through /proc (see later section).
113 AFS has a concept of mountpoints. In AFS terms, these are specially formatted
114 symbolic links (of the same form as the "device name" passed to mount). kAFS
115 presents these to the user as directories that have a follow-link capability
116 (ie: symbolic link semantics). If anyone attempts to access them, they will
117 automatically cause the target volume to be mounted (if possible) on that site.
119 Automatically mounted filesystems will be automatically unmounted approximately
120 twenty minutes after they were last used. Alternatively they can be unmounted
121 directly with the umount() system call.
123 Manually unmounting an AFS volume will cause any idle submounts upon it to be
124 culled first. If all are culled, then the requested volume will also be
125 unmounted, otherwise error EBUSY will be returned.
127 This can be used by the administrator to attempt to unmount the whole AFS tree
128 mounted on /afs in one go by doing:
137 The AFS modules creates a "/proc/fs/afs/" directory and populates it:
139 (*) A "cells" file that lists cells currently known to the afs module and
142 [root@andromeda ~]# cat /proc/fs/afs/cells
144 3 cambridge.redhat.com
146 (*) A directory per cell that contains files that list volume location
147 servers, volumes, and active servers known within that cell.
149 [root@andromeda ~]# cat /proc/fs/afs/cambridge.redhat.com/servers
152 [root@andromeda ~]# cat /proc/fs/afs/cambridge.redhat.com/vlservers
155 [root@andromeda ~]# cat /proc/fs/afs/cambridge.redhat.com/volumes
156 USE STT VLID[0] VLID[1] VLID[2] NAME
157 1 Val 20000000 20000001 20000002 root.afs
164 The filesystem maintains an internal database of all the cells it knows and the
165 IP addresses of the volume location servers for those cells. The cell to which
166 the system belongs is added to the database when insmod is performed by the
167 "rootcell=" argument or, if compiled in, using a "kafs.rootcell=" argument on
168 the kernel command line.
170 Further cells can be added by commands similar to the following:
172 echo add CELLNAME VLADDR[:VLADDR][:VLADDR]... >/proc/fs/afs/cells
173 echo add grand.central.org 18.7.14.88:128.2.191.224 >/proc/fs/afs/cells
175 No other cell database operations are available at this time.
182 Secure operations are initiated by acquiring a key using the klog program. A
183 very primitive klog program is available at:
185 http://people.redhat.com/~dhowells/rxrpc/klog.c
187 This should be compiled by:
189 make klog LDLIBS="-lcrypto -lcrypt -lkrb4 -lkeyutils"
195 Assuming it's successful, this adds a key of type RxRPC, named for the service
196 and cell, eg: "afs@<cellname>". This can be viewed with the keyctl program or
197 by cat'ing /proc/keys:
199 [root@andromeda ~]# keyctl show
201 -3 --alswrv 0 0 keyring: _ses.3268
202 2 --alswrv 0 0 \_ keyring: _uid.0
203 111416553 --als--v 0 0 \_ rxrpc: afs@CAMBRIDGE.REDHAT.COM
205 Currently the username, realm, password and proposed ticket lifetime are
206 compiled in to the program.
208 It is not required to acquire a key before using AFS facilities, but if one is
209 not acquired then all operations will be governed by the anonymous user parts
212 If a key is acquired, then all AFS operations, including mounts and automounts,
213 made by a possessor of that key will be secured with that key.
215 If a file is opened with a particular key and then the file descriptor is
216 passed to a process that doesn't have that key (perhaps over an AF_UNIX
217 socket), then the operations on the file will be made with key that was used to
225 Here's what I use to test this. Some of the names and IP addresses are local
226 to my internal DNS. My "root.afs" partition has a mount point within it for
227 some public volumes volumes.
231 insmod /tmp/kafs.o rootcell=cambridge.redhat.com:172.16.18.91
233 mount -t afs \%root.afs. /afs
234 mount -t afs \%cambridge.redhat.com:root.cell. /afs/cambridge.redhat.com/
236 echo add grand.central.org 18.7.14.88:128.2.191.224 > /proc/fs/afs/cells
237 mount -t afs "#grand.central.org:root.cell." /afs/grand.central.org/
238 mount -t afs "#grand.central.org:root.archive." /afs/grand.central.org/archive
239 mount -t afs "#grand.central.org:root.contrib." /afs/grand.central.org/contrib
240 mount -t afs "#grand.central.org:root.doc." /afs/grand.central.org/doc
241 mount -t afs "#grand.central.org:root.project." /afs/grand.central.org/project
242 mount -t afs "#grand.central.org:root.service." /afs/grand.central.org/service
243 mount -t afs "#grand.central.org:root.software." /afs/grand.central.org/software
244 mount -t afs "#grand.central.org:root.user." /afs/grand.central.org/user