freebsd-nq/usr.sbin/jail/jail.8

1428 lines
42 KiB
Groff
Raw Normal View History

.\" Copyright (c) 2000, 2003 Robert N. M. Watson
.\" Copyright (c) 2008-2012 James Gritton
2001-12-14 10:18:15 +00:00
.\" All rights reserved.
.\"
2001-12-14 10:18:15 +00:00
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
2001-12-14 10:18:15 +00:00
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
2001-12-14 10:18:15 +00:00
.\" $FreeBSD$
.\"
.Dd November 18, 2020
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Dt JAIL 8
.Os
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Sh NAME
.Nm jail
.Nd "manage system jails"
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Sh SYNOPSIS
.Nm
.Op Fl dhilqv
.Op Fl J Ar jid_file
.Op Fl u Ar username
.Op Fl U Ar username
.Op Fl cmr
.Ar param Ns = Ns Ar value ...
.Op Cm command Ns = Ns Ar command ...
.Nm
.Op Fl dqv
.Op Fl f Ar conf_file
.Op Fl p Ar limit
.Op Fl cmr
.Op Ar jail
.Nm
.Op Fl qv
.Op Fl f Ar conf_file
.Op Fl rR
.Op Cm * | Ar jail ...
.Nm
.Op Fl dhilqv
.Op Fl J Ar jid_file
.Op Fl u Ar username
.Op Fl U Ar username
.Op Fl n Ar jailname
.Op Fl s Ar securelevel
.Op Ar path hostname [ Ar ip Ns [ Ns Ar ,... Ns ]] Ar command ...
.Nm
.Op Fl f Ar conf_file
.Fl e
.Ar separator
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Sh DESCRIPTION
The
.Nm
utility creates new jails, or modifies or removes existing jails.
It can also print a list of configured jails and their parameters.
A jail
.Pq or Dq prison
is specified via parameters on the command line, or in the
.Xr jail.conf 5
file.
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Pp
At least one of the options
.Fl c ,
.Fl e ,
.Fl m
or
.Fl r
must be specified.
These options are used alone or in combination to describe the operation to
perform:
.Bl -tag -width indent
.It Fl c
Create a new jail.
2012-05-12 19:59:37 +00:00
The jail
.Va jid
and
.Va name
parameters (if specified on the command line)
must not refer to an existing jail.
.It Fl e Ar separator
Exhibit a list of all configured non-wildcard jails and their parameters.
No jail creation, modification or removal performed if this option is used.
The
.Ar separator
string is used to separate parameters.
Use
.Xr jls 8
utility to list running jails.
.It Fl m
Modify an existing jail.
One of the
.Va jid
or
.Va name
parameters must exist and refer to an existing jail.
Some parameters may not be changed on a running jail.
.It Fl r
Remove the
.Ar jail
specified by jid or name.
All jailed processes are killed, and all jails that are
children of this jail are also
removed.
.It Fl rc
Restart an existing jail.
The jail is first removed and then re-created, as if
.Dq Nm Fl r
and
.Dq Nm Fl c
were run in succession.
.It Fl cm
Create a jail if it does not exist, or modify the jail if it does exist.
.It Fl mr
Modify an existing jail.
The jail may be restarted if necessary to modify parameters than could
not otherwise be changed.
.It Fl cmr
Create a jail if it doesn't exist, or modify (and possibly restart) the
jail if it does exist.
.El
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Pp
Other available options are:
.Bl -tag -width indent
.It Fl d
Allow making changes to a dying jail, equivalent to the
.Va allow.dying
parameter.
.It Fl f Ar conf_file
Use configuration file
.Ar conf_file
instead of the default
.Pa /etc/jail.conf .
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
.It Fl h
Resolve the
.Va host.hostname
parameter (or
.Va hostname )
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
and add all IP addresses returned by the resolver
to the list of addresses for this jail.
This is equivalent to the
.Va ip_hostname
parameter.
.It Fl i
Output (only) the jail identifier of the newly created jail(s).
This implies the
.Fl q
option.
.It Fl J Ar jid_file
2006-09-29 17:57:04 +00:00
Write a
.Ar jid_file
file, containing the parameters used to start the jail.
.It Fl l
Run commands in a clean environment.
This is deprecated and is equivalent to the exec.clean parameter.
.It Fl n Ar jailname
Set the jail's name.
This is deprecated and is equivalent to the
.Va name
parameter.
.It Fl p Ar limit
Limit the number of commands from
.Va exec.*
that can run simultaneously.
.It Fl q
Suppress the message printed whenever a jail is created, modified or removed.
Only error messages will be printed.
.It Fl R
A variation of the
.Fl r
option that removes an existing jail without using the configuration file.
No removal-related parameters for this jail will be used \(em the jail will
simply be removed.
.It Fl s Ar securelevel
Set the
2006-09-29 17:57:04 +00:00
.Va kern.securelevel
MIB entry to the specified value inside the newly created jail.
This is deprecated and is equivalent to the
.Va securelevel
parameter.
.It Fl u Ar username
The user name from host environment as whom jailed commands should run.
This is deprecated and is equivalent to the
.Va exec.jail_user
and
.Va exec.system_jail_user
parameters.
.It Fl U Ar username
The user name from the jailed environment as whom jailed commands should run.
This is deprecated and is equivalent to the
.Va exec.jail_user
parameter.
.It Fl v
Print a message on every operation, such as running commands and
mounting filesystems.
.El
.Pp
If no arguments are given after the options, the operation (except
remove) will be performed on all jails specified in the
.Xr jail.conf 5
file.
A single argument of a jail name will operate only on the specified jail.
The
.Fl r
and
.Fl R
options can also remove running jails that aren't in the
.Xr jail.conf 5
file, specified by name or jid.
2012-05-11 22:05:30 +00:00
.Pp
An argument of
.Dq *
is a wildcard that will operate on all jails, regardless of whether
they appear in
.Xr jail.conf 5 ;
this is the surest way for
.Fl r
to remove all jails.
If hierarchical jails exist, a partial-matching wildcard definition may
be specified.
2012-05-12 19:59:37 +00:00
For example, an argument of
.Dq foo.*
would apply to jails with names like
.Dq foo.bar
and
.Dq foo.bar.baz .
.Pp
A jail may be specified with parameters directly on the command line.
In this case, the
.Xr jail.conf 5
file will not be used.
For backward compatibility, the command line may also have four fixed
parameters, without names:
.Ar path ,
.Ar hostname ,
.Ar ip ,
and
.Ar command .
This mode will always create a new jail, and the
.Fl c
and
.Fl m
options do not apply (and must not be present).
.Ss Jail Parameters
Parameters in the
.Xr jail.conf 5
file, or on the command line, are generally of the form
.Dq name=value .
Some parameters are boolean, and do not have a value but are set by the
name alone with or without a
.Dq no
prefix, e.g.
.Va persist
or
.Va nopersist .
They can also be given the values
.Dq true
and
.Dq false .
Other parameters may have more than one value, specified as a
comma-separated list or with
.Dq +=
in the configuration file (see
.Xr jail.conf 5
for details).
.Pp
The
.Nm
utility recognizes two classes of parameters.
There are the true jail
parameters that are passed to the kernel when the jail is created,
which can be seen with
.Xr jls 8 ,
and can (usually) be changed with
.Dq Nm Fl m .
Then there are pseudo-parameters that are only used by
.Nm
itself.
.Pp
Jails have a set of core parameters, and kernel modules can add their own
jail parameters.
The current set of available parameters can be retrieved via
.Dq Nm sysctl Fl d Va security.jail.param .
Any parameters not set will be given default values, often based on the
current environment.
The core parameters are:
.Bl -tag -width indent
.It Va jid
The jail identifier.
This will be assigned automatically to a new jail (or can be explicitly
set), and can be used to identify the jail for later modification, or
for such commands as
.Xr jls 8
or
.Xr jexec 8 .
.It Va name
The jail name.
This is an arbitrary string that identifies a jail (except it may not
contain a
.Sq \&. ) .
Like the
.Va jid ,
it can be passed to later
.Nm
commands, or to
.Xr jls 8
or
.Xr jexec 8 .
If no
.Va name
is supplied, a default is assumed that is the same as the
.Va jid .
The
.Va name
parameter is implied by the
.Xr jail.conf 5
file format, and need not be explicitly set when using the configuration
file.
.It Va path
The directory which is to be the root of the jail.
Any commands run inside the jail, either by
.Nm
or from
.Xr jexec 8 ,
are run from this directory.
.It Va ip4.addr
A list of IPv4 addresses assigned to the jail.
2010-05-05 08:43:47 +00:00
If this is set, the jail is restricted to using only these addresses.
Any attempts to use other addresses fail, and attempts to use wildcard
addresses silently use the jailed address instead.
For IPv4 the first address given will be used as the source address
when source address selection on unbound sockets cannot find a better
match.
It is only possible to start multiple jails with the same IP address
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
if none of the jails has more than this single overlapping IP address
assigned to itself.
.It Va ip4.saddrsel
A boolean option to change the formerly mentioned behaviour and disable
IPv4 source address selection for the jail in favour of the primary
IPv4 address of the jail.
Source address selection is enabled by default for all jails and the
.Va ip4.nosaddrsel
setting of a parent jail is not inherited for any child jails.
.It Va ip4
2010-08-01 09:37:36 +00:00
Control the availability of IPv4 addresses.
Possible values are
.Dq inherit
to allow unrestricted access to all system addresses,
.Dq new
to restrict addresses via
.Va ip4.addr ,
and
.Dq disable
to stop the jail from using IPv4 entirely.
Setting the
.Va ip4.addr
parameter implies a value of
.Dq new .
.It Va ip6.addr , Va ip6.saddrsel , Va ip6
A set of IPv6 options for the jail, the counterparts to
.Va ip4.addr ,
.Va ip4.saddrsel
and
.Va ip4
above.
.It Va vnet
Create the jail with its own virtual network stack,
with its own network interfaces, addresses, routing table, etc.
The kernel must have been compiled with the
.Sy VIMAGE option
for this to be available.
Possible values are
.Dq inherit
to use the system network stack, possibly with restricted IP addresses,
and
.Dq new
to create a new network stack.
.It Va host.hostname
The hostname of the jail.
Other similar parameters are
.Va host.domainname ,
.Va host.hostuuid
and
.Va host.hostid .
.It Va host
Set the origin of hostname and related information.
Possible values are
.Dq inherit
to use the system information and
.Dq new
for the jail to use the information from the above fields.
Setting any of the above fields implies a value of
.Dq new .
.It Va securelevel
The value of the jail's
.Va kern.securelevel
sysctl.
A jail never has a lower securelevel than its parent system, but by
setting this parameter it may have a higher one.
If the system securelevel is changed, any jail securelevels will be at
least as secure.
.It Va devfs_ruleset
The number of the devfs ruleset that is enforced for mounting devfs in
this jail.
A value of zero (default) means no ruleset is enforced.
Descendant jails inherit the parent jail's devfs ruleset enforcement.
Mounting devfs inside a jail is possible only if the
.Va allow.mount
and
.Va allow.mount.devfs
permissions are effective and
.Va enforce_statfs
is set to a value lower than 2.
Devfs rules and rulesets cannot be viewed or modified from inside a jail.
.Pp
NOTE: It is important that only appropriate device nodes in devfs be
exposed to a jail; access to disk devices in the jail may permit processes
in the jail to bypass the jail sandboxing by modifying files outside of
the jail.
See
.Xr devfs 8
for information on how to use devfs rules to limit access to entries
in the per-jail devfs.
A simple devfs ruleset for jails is available as ruleset #4 in
.Pa /etc/defaults/devfs.rules .
.It Va children.max
The number of child jails allowed to be created by this jail (or by
other jails under this jail).
This limit is zero by default, indicating the jail is not allowed to
create child jails.
See the
.Sx "Hierarchical Jails"
section for more information.
.It Va children.cur
2012-06-03 11:29:48 +00:00
The number of descendants of this jail, including its own child jails
and any jails created under them.
.It Va enforce_statfs
This determines what information processes in a jail are able to get
about mount points.
It affects the behaviour of the following syscalls:
.Xr statfs 2 ,
.Xr fstatfs 2 ,
.Xr getfsstat 2 ,
and
.Xr fhstatfs 2
(as well as similar compatibility syscalls).
When set to 0, all mount points are available without any restrictions.
When set to 1, only mount points below the jail's chroot directory are
visible.
In addition to that, the path to the jail's chroot directory is removed
from the front of their pathnames.
When set to 2 (default), above syscalls can operate only on a mount-point
where the jail's chroot directory is located.
.It Va persist
Setting this boolean parameter allows a jail to exist without any
processes.
Normally, a command is run as part of jail creation, and then the jail
is destroyed as its last process exits.
A new jail must have either the
.Va persist
parameter or
.Va exec.start
or
.Va command
pseudo-parameter set.
.It Va cpuset.id
The ID of the cpuset associated with this jail (read-only).
.It Va dying
This is true if the jail is in the process of shutting down (read-only).
.It Va parent
The
.Va jid
of the parent of this jail, or zero if this is a top-level jail
(read-only).
.It Va osrelease
The string for the jail's
.Va kern.osrelease
sysctl and uname -r.
.It Va osreldate
The number for the jail's
.Va kern.osreldate
and uname -K.
.It Va allow.*
Some restrictions of the jail environment may be set on a per-jail
basis.
With the exception of
.Va allow.set_hostname
and
.Va allow.reserved_ports ,
these boolean parameters are off by default.
.Bl -tag -width indent
.It Va allow.set_hostname
The jail's hostname may be changed via
.Xr hostname 1
or
.Xr sethostname 3 .
.It Va allow.sysvipc
A process within the jail has access to System V IPC primitives.
This is deprecated in favor of the per-module parameters (see below).
When this parameter is set, it is equivalent to setting
.Va sysvmsg ,
.Va sysvsem ,
and
.Va sysvshm
all to
.Dq inherit .
.It Va allow.raw_sockets
The jail root is allowed to create raw sockets.
Setting this parameter allows utilities like
.Xr ping 8
and
.Xr traceroute 8
to operate inside the jail.
If this is set, the source IP addresses are enforced to comply
with the IP address bound to the jail, regardless of whether or not
the
.Dv IP_HDRINCL
flag has been set on the socket.
Since raw sockets can be used to configure and interact with various
network subsystems, extra caution should be used where privileged access
to jails is given out to untrusted parties.
.It Va allow.chflags
Normally, privileged users inside a jail are treated as unprivileged by
.Xr chflags 2 .
When this parameter is set, such users are treated as privileged, and
may manipulate system file flags subject to the usual constraints on
.Va kern.securelevel .
.It Va allow.mount
privileged users inside the jail will be able to mount and unmount file
system types marked as jail-friendly.
The
.Xr lsvfs 1
command can be used to find file system types available for mount from
within a jail.
This permission is effective only if
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.devfs
privileged users inside the jail will be able to mount and unmount the
devfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
The devfs ruleset should be restricted from the default by using the
.Va devfs_ruleset
option.
.It Va allow.quotas
The jail root may administer quotas on the jail's filesystem(s).
This includes filesystems that the jail may share with other jails or
with non-jailed parts of the system.
.It Va allow.read_msgbuf
Jailed users may read the kernel message buffer.
If the
.Va security.bsd.unprivileged_read_msgbuf
MIB entry is zero, this will be restricted to the root user.
.It Va allow.socket_af
Sockets within a jail are normally restricted to IPv4, IPv6, local
(UNIX), and route. This allows access to other protocol stacks that
have not had jail functionality added to them.
.It Va allow.mlock
Locking or unlocking physical pages in memory are normally not available
within a jail.
When this parameter is set, users may
.Xr mlock 2
or
.Xr munlock 2
memory subject to
.Va security.bsd.unprivileged_mlock
and resource limits.
.It Va allow.reserved_ports
The jail root may bind to ports lower than 1024.
2018-11-27 17:51:50 +00:00
.It Va allow.unprivileged_proc_debug
Unprivileged processes in the jail may use debugging facilities.
.It Va allow.suser
The value of the jail's
.Va security.bsd.suser_enabled
sysctl.
The super-user will be disabled automatically if its parent system has it
disabled.
The super-user is enabled by default.
.El
.El
.Pp
Kernel modules may add their own parameters, which only exist when the
module is loaded.
These are typically headed under a parameter named after the module,
with values of
.Dq inherit
to give the jail full use of the module,
.Dq new
to encapsulate the jail in some module-specific way,
and
.Dq disable
to make the module unavailable to the jail.
There also may be other parameters to define jail behavior within the module.
Module-specific parameters include:
.Bl -tag -width indent
.It Va allow.mount.fdescfs
privileged users inside the jail will be able to mount and unmount the
fdescfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.fusefs
privileged users inside the jail will be able to mount and unmount
fuse-based file systems.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.nullfs
privileged users inside the jail will be able to mount and unmount the
nullfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.procfs
privileged users inside the jail will be able to mount and unmount the
procfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.linprocfs
privileged users inside the jail will be able to mount and unmount the
linprocfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.linsysfs
privileged users inside the jail will be able to mount and unmount the
linsysfs file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
2013-08-23 22:52:20 +00:00
.It Va allow.mount.tmpfs
privileged users inside the jail will be able to mount and unmount the
tmpfs file system.
This permission is effective only together with
.Va allow.mount
and only when
2013-08-23 22:52:20 +00:00
.Va enforce_statfs
is set to a value lower than 2.
.It Va allow.mount.zfs
privileged users inside the jail will be able to mount and unmount the
ZFS file system.
This permission is effective only together with
.Va allow.mount
and only when
.Va enforce_statfs
is set to a value lower than 2.
See
.Xr zfs 8
for information on how to configure the ZFS filesystem to operate from
within a jail.
.It Va allow.vmm
The jail may access
.Xr vmm 4 .
This flag is only available when the
.Xr vmm 4
kernel module is loaded.
.It Va linux
Determine how a jail's Linux emulation environment appears.
A value of
.Dq inherit
will keep the same environment, and
.Dq new
add ability to set watchdog timeout for a shutdown This change allows to specify a watchdog(9) timeout for a system shutdown. The timeout is activated when the watchdogd daemon is stopped. The idea is to a prevent any indefinite hang during late stages of the shutdown. The feature is implemented in rc.d/watchdogd, it builds upon watchdogd -x option. Note that the shutdown timeout is not actiavted when the watchdogd service is individually stopped by an operator. It is also not activated for the 'shutdown' to the single-user mode. In those cases it is assumed that the operator knows what they are doing and they have means to recover the system should it hang. Significant subchanges and implementation details: - the argument to rc.shutdown, completely unused before, is assigned to rc_shutdown variable that can be inspected by rc scripts - init(8) passes "single" or "reboot" as the argument, this is not changed - the argument is not mandatory and if it is not set then rc_shutdown is set to "unspecified" - however, the default jail management scripts and jail configuration examples have been updated to pass "jail" to rc.shutdown, just in case - the new timeout can be set via watchdogd_shutdown_timeout rc option - for consistency, the regular timeout can now be set via watchdogd_timeout rc option - watchdogd_shutdown_timeout and watchdogd_timeout override timeout specifications in watchdogd_flags - existing configurations, where the new rc options are not set, should keep working as before I am not particularly wed to any of the implementation specifics. I am open to changing or removing any of them as long as the provided functionality is the same (or very close) to the proposed one. For example, I think it can be implemented without using watchdogd -x, by means of watchdog(1) alone. In that case there would be a small window between stopping watchdogd and running watchdog, but I think that that is acceptable. Reviewed by: bcr (man page changes) MFC after: 5 weeks Relnotes: yes Differential Revision: https://reviews.freebsd.org/D21221
2019-10-03 11:23:10 +00:00
will give the jail its own environment (still originally inherited when
the jail is created).
.It Va linux.osname , linux.osrelease , linux.oss_version
The Linux OS name, OS release, and OSS version associated with this jail.
.It Va sysvmsg
Allow access to SYSV IPC message primitives.
If set to
.Dq inherit ,
all IPC objects on the system are visible to this jail, whether they
were created by the jail itself, the base system, or other jails.
If set to
.Dq new ,
the jail will have its own key namespace, and can only see the objects
that it has created;
the system (or parent jail) has access to the jail's objects, but not to
its keys.
If set to
.Dq disable ,
the jail cannot perform any sysvmsg-related system calls.
2016-05-01 16:48:03 +00:00
.It Va sysvsem, sysvshm
Allow access to SYSV IPC semaphore and shared memory primitives, in the
same manner as
.Va sysvmsg.
.El
.Pp
There are pseudo-parameters that are not passed to the kernel, but are
used by
.Nm
to set up the jail environment, often by running specified commands
when jails are created or removed.
The
.Va exec.*
command parameters are
.Xr sh 1
command lines that are run in either the system or jail environment.
They may be given multiple values, which would run the specified
commands in sequence.
All commands must succeed (return a zero exit status), or the jail will
not be created or removed, as appropriate.
.Pp
The pseudo-parameters are:
.Bl -tag -width indent
.It Va exec.prepare
Command(s) to run in the system environment to prepare a jail for creation.
These commands are executed before assigning IP addresses and mounting
filesystems, so they may be used to create a new jail filesystem if it does
not already exist.
.It Va exec.prestart
Command(s) to run in the system environment before a jail is created.
.It Va exec.created
Command(s) to run in the system environment right after a jail has been
created, but before commands (or services) get executed in the jail.
.It Va exec.start
Command(s) to run in the jail environment when a jail is created.
A typical command to run is
.Dq sh /etc/rc .
.It Va command
A synonym for
.Va exec.start
for use when specifying a jail directly on the command line.
Unlike other parameters whose value is a single string,
.Va command
uses the remainder of the
.Nm
command line as its own arguments.
.It Va exec.poststart
Command(s) to run in the system environment after a jail is created,
and after any
.Va exec.start
commands have completed.
.It Va exec.prestop
Command(s) to run in the system environment before a jail is removed.
.It Va exec.stop
Command(s) to run in the jail environment before a jail is removed,
and after any
.Va exec.prestop
commands have completed.
A typical command to run is
add ability to set watchdog timeout for a shutdown This change allows to specify a watchdog(9) timeout for a system shutdown. The timeout is activated when the watchdogd daemon is stopped. The idea is to a prevent any indefinite hang during late stages of the shutdown. The feature is implemented in rc.d/watchdogd, it builds upon watchdogd -x option. Note that the shutdown timeout is not actiavted when the watchdogd service is individually stopped by an operator. It is also not activated for the 'shutdown' to the single-user mode. In those cases it is assumed that the operator knows what they are doing and they have means to recover the system should it hang. Significant subchanges and implementation details: - the argument to rc.shutdown, completely unused before, is assigned to rc_shutdown variable that can be inspected by rc scripts - init(8) passes "single" or "reboot" as the argument, this is not changed - the argument is not mandatory and if it is not set then rc_shutdown is set to "unspecified" - however, the default jail management scripts and jail configuration examples have been updated to pass "jail" to rc.shutdown, just in case - the new timeout can be set via watchdogd_shutdown_timeout rc option - for consistency, the regular timeout can now be set via watchdogd_timeout rc option - watchdogd_shutdown_timeout and watchdogd_timeout override timeout specifications in watchdogd_flags - existing configurations, where the new rc options are not set, should keep working as before I am not particularly wed to any of the implementation specifics. I am open to changing or removing any of them as long as the provided functionality is the same (or very close) to the proposed one. For example, I think it can be implemented without using watchdogd -x, by means of watchdog(1) alone. In that case there would be a small window between stopping watchdogd and running watchdog, but I think that that is acceptable. Reviewed by: bcr (man page changes) MFC after: 5 weeks Relnotes: yes Differential Revision: https://reviews.freebsd.org/D21221
2019-10-03 11:23:10 +00:00
.Dq sh /etc/rc.shutdown jail .
.It Va exec.poststop
Command(s) to run in the system environment after a jail is removed.
.It Va exec.release
Command(s) to run in the system environment after all other actions are done.
These commands are executed after unmounting filesystems and removing IP
addresses, so they may be used to remove a jail filesystem if it is no longer
needed.
.It Va exec.clean
Run commands in a clean environment.
The environment is discarded except for
.Ev HOME , SHELL , TERM
and
.Ev USER .
.Ev HOME
and
.Ev SHELL
are set to the target login's default values.
.Ev USER
is set to the target login.
.Ev TERM
is imported from the current environment.
The environment variables from the login class capability database for the
target login are also set.
.It Va exec.jail_user
The user to run commands as, when running in the jail environment.
The default is to run the commands as the current user.
.It Va exec.system_jail_user
This boolean option looks for the
.Va exec.jail_user
in the system
.Xr passwd 5
file, instead of in the jail's file.
.It Va exec.system_user
The user to run commands as, when running in the system environment.
The default is to run the commands as the current user.
.It Va exec.timeout
The maximum amount of time to wait for a command to complete, in
seconds.
If a command is still running after this timeout has passed,
the jail will not be created or removed, as appropriate.
.It Va exec.consolelog
A file to direct command output (stdout and stderr) to.
.It Va exec.fib
The FIB (routing table) to set when running commands inside the jail.
.It Va stop.timeout
The maximum amount of time to wait for a jail's processes to exit
after sending them a
.Dv SIGTERM
signal (which happens after the
.Va exec.stop
commands have completed).
After this many seconds have passed, the jail will be removed, which
will kill any remaining processes.
If this is set to zero, no
.Dv SIGTERM
is sent and the jail is immediately removed.
The default is 10 seconds.
.It Va interface
A network interface to add the jail's IP addresses
.Va ( ip4.addr
and
.Va ip6.addr )
to.
An alias for each address will be added to the interface before the
jail is created, and will be removed from the interface after the
jail is removed.
.It Va ip4.addr
In addition to the IP addresses that are passed to the kernel, an
interface, netmask and additional parameters (as supported by
.Xr ifconfig 8 Ns )
may also be specified, in the form
.Dq Ar interface Ns | Ns Ar ip-address Ns / Ns Ar netmask param ... .
If an interface is given before the IP address, an alias for the address
will be added to that interface, as it is with the
.Va interface
parameter.
If a netmask in either dotted-quad or CIDR form is given
after an IP address, it will be used when adding the IP alias.
If additional parameters are specified then they will also be used when
adding the IP alias.
.It Va ip6.addr
In addition to the IP addresses that are passed to the kernel,
an interface, prefix and additional parameters (as supported by
.Xr ifconfig 8 Ns )
may also be specified, in the form
.Dq Ar interface Ns | Ns Ar ip-address Ns / Ns Ar prefix param ... .
.It Va vnet.interface
A network interface to give to a vnet-enabled jail after is it created.
The interface will automatically be released when the jail is removed.
.It Va ip_hostname
Resolve the
.Va host.hostname
parameter and add all IP addresses returned by the resolver
to the list of addresses
.Po Va ip4.addr
or
.Va ip6.addr Pc
for this jail.
This may affect default address selection for outgoing IPv4 connections
from jails.
The address first returned by the resolver for each address family
will be used as the primary address.
.It Va mount
A filesystem to mount before creating the jail (and to unmount after
removing it), given as a single
.Xr fstab 5
line.
.It Va mount.fstab
An
.Xr fstab 5
format file containing filesystems to mount before creating a jail.
.It Va mount.devfs
Mount a
.Xr devfs 5
filesystem on the chrooted
.Pa /dev
directory, and apply the ruleset in the
.Va devfs_ruleset
parameter (or a default of ruleset 4: devfsrules_jail)
to restrict the devices visible inside the jail.
.It Va mount.fdescfs
Mount a
.Xr fdescfs 5
filesystem on the chrooted
.Pa /dev/fd
directory.
.It Va mount.procfs
Mount a
.Xr procfs 5
filesystem on the chrooted
.Pa /proc
directory.
.It Va allow.dying
Allow making changes to a
.Va dying
jail.
.It Va depend
Specify a jail (or jails) that this jail depends on.
When this jail is to be created, any jail(s) it depends on must already exist.
If not, they will be created automatically, up to the completion of the last
.Va exec.poststart
command, before any action will taken to create this jail.
When jails are removed the opposite is true:
this jail will be removed, up to the last
.Va exec.poststop
command, before any jail(s) it depends on are stopped.
.El
.Sh EXAMPLES
Jails are typically set up using one of two philosophies: either to
constrain a specific application (possibly running with privilege), or
2004-06-05 20:27:10 +00:00
to create a
.Dq "virtual system image"
running a variety of daemons and services.
In both cases, a fairly complete file system install of
.Fx
is
required, so as to provide the necessary command line tools, daemons,
2004-05-20 06:37:44 +00:00
libraries, application configuration files, etc.
However, for a virtual server configuration, a fair amount of
additional work is required so as to replace the
2004-06-05 20:27:10 +00:00
.Dq boot
process.
2004-05-20 06:37:44 +00:00
This manual page documents the configuration steps necessary to support
either of these steps, although the configuration steps may need to be
refined based on local requirements.
2001-12-14 10:18:15 +00:00
.Ss "Setting up a Jail Directory Tree"
To set up a jail directory tree containing an entire
.Fx
distribution, the following
.Xr sh 1
command script can be used:
2001-07-15 08:06:20 +00:00
.Bd -literal
1999-07-09 21:35:50 +00:00
D=/here/is/the/jail
cd /usr/src
mkdir -p $D
make world DESTDIR=$D
make distribution DESTDIR=$D
1999-07-09 21:35:50 +00:00
.Ed
.Pp
2004-05-20 06:37:44 +00:00
In many cases this example would put far more in the jail than needed.
In the other extreme case a jail might contain only one file:
the executable to be run in the jail.
.Pp
We recommend experimentation, and caution that it is a lot easier to
start with a
.Dq fat
jail and remove things until it stops working,
than it is to start with a
.Dq thin
jail and add things until it works.
2001-12-14 10:18:15 +00:00
.Ss "Setting Up a Jail"
Do what was described in
2001-12-14 10:18:15 +00:00
.Sx "Setting Up a Jail Directory Tree"
to build the jail directory tree.
For the sake of this example, we will
assume you built it in
.Pa /data/jail/testjail ,
for a jail named
.Dq testjail .
2001-12-14 10:18:15 +00:00
Substitute below as needed with your
own directory, IP address, and hostname.
.Ss "Setting up the Host Environment"
First, set up the real system's environment to be
.Dq jail-friendly .
For consistency, we will refer to the parent box as the
2001-12-14 10:18:15 +00:00
.Dq "host environment" ,
and to the jailed virtual machine as the
2001-12-14 10:18:15 +00:00
.Dq "jail environment" .
Since jails are implemented using IP aliases, one of the first things to do
is to disable IP services on the host system that listen on all local
2001-12-14 10:18:15 +00:00
IP addresses for a service.
If a network service is present in the host environment that binds all
available IP addresses rather than specific IP addresses, it may service
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
requests sent to jail IP addresses if the jail did not bind the port.
2001-12-14 10:18:15 +00:00
This means changing
.Xr inetd 8
to only listen on the
2001-12-14 10:18:15 +00:00
appropriate IP address, and so forth.
Add the following to
.Pa /etc/rc.conf
in the host environment:
.Bd -literal -offset indent
sendmail_enable="NO"
inetd_flags="-wW -a 192.0.2.23"
rpcbind_enable="NO"
.Ed
.Pp
.Li 192.0.2.23
2001-12-14 10:18:15 +00:00
is the native IP address for the host system, in this example.
Daemons that run out of
.Xr inetd 8
can be easily configured to use only the specified host IP address.
2001-12-14 10:18:15 +00:00
Other daemons
will need to be manually configured \(em for some this is possible through
.Xr rc.conf 5
2004-05-20 06:37:44 +00:00
flags entries; for others it is necessary to modify per-application
configuration files, or to recompile the application.
The following frequently deployed services must have their individual
configuration files modified to limit the application to listening
to a specific IP address:
.Pp
To configure
.Xr sshd 8 ,
it is necessary to modify
.Pa /etc/ssh/sshd_config .
.Pp
To configure
.Xr sendmail 8 ,
it is necessary to modify
.Pa /etc/mail/sendmail.cf .
.Pp
For
.Xr named 8 ,
it is necessary to modify
.Pa /etc/namedb/named.conf .
.Pp
In addition, a number of services must be recompiled in order to run
them in the host environment.
This includes most applications providing services using
.Xr rpc 3 ,
such as
2005-01-21 20:48:00 +00:00
.Xr rpcbind 8 ,
.Xr nfsd 8 ,
and
.Xr mountd 8 .
In general, applications for which it is not possible to specify which
IP address to bind should not be run in the host environment unless they
should also service requests sent to jail IP addresses.
Attempting to serve
NFS from the host environment may also cause confusion, and cannot be
easily reconfigured to use only specific IPs, as some NFS services are
2001-12-14 10:18:15 +00:00
hosted directly from the kernel.
2004-05-20 06:37:44 +00:00
Any third-party network software running
in the host environment should also be checked and configured so that it
does not bind all IP addresses, which would result in those services also
appearing to be offered by the jail environments.
.Pp
Once
these daemons have been disabled or fixed in the host environment, it is
best to reboot so that all daemons are in a known state, to reduce the
potential for confusion later (such as finding that when you send mail
to a jail, and its sendmail is down, the mail is delivered to the host,
2004-06-05 20:27:10 +00:00
etc.).
.Ss "Configuring the Jail"
2004-05-20 06:37:44 +00:00
Start any jail for the first time without configuring the network
2001-12-14 10:18:15 +00:00
interface so that you can clean it up a little and set up accounts.
As
with any machine (virtual or not), you will need to set a root password, time
2001-12-14 10:18:15 +00:00
zone, etc.
Some of these steps apply only if you intend to run a full virtual server
2004-05-20 06:37:44 +00:00
inside the jail; others apply both for constraining a particular application
or for running a virtual server.
.Pp
Start a shell in the jail:
.Bd -literal -offset indent
jail -c path=/data/jail/testjail mount.devfs \\
host.hostname=testhostname ip4.addr=192.0.2.100 \\
command=/bin/sh
.Ed
2001-12-14 10:18:15 +00:00
.Pp
2004-05-20 06:37:44 +00:00
Assuming no errors, you will end up with a shell prompt within the jail.
2001-12-14 10:18:15 +00:00
You can now run
.Xr bsdconfig 8
and do the post-install configuration to set various configuration options,
2001-12-14 10:18:15 +00:00
or perform these actions manually by editing
.Pa /etc/rc.conf ,
etc.
.Pp
.Bl -bullet -offset indent -compact
.It
Configure
.Pa /etc/resolv.conf
so that name resolution within the jail will work correctly.
.It
Run
.Xr newaliases 1
2001-12-14 10:18:15 +00:00
to quell
.Xr sendmail 8
warnings.
.It
Set a root password, probably different from the real host system.
.It
Set the timezone.
.It
Add accounts for users in the jail environment.
.It
Install any packages the environment requires.
.El
.Pp
You may also want to perform any package-specific configuration (web servers,
2001-12-14 10:18:15 +00:00
SSH servers, etc), patch up
.Pa /etc/syslog.conf
so it logs as you would like, etc.
If you are not using a virtual server, you may wish to modify
.Xr syslogd 8
in the host environment to listen on the syslog socket in the jail
environment; in this example, the syslog socket would be stored in
.Pa /data/jail/testjail/var/run/log .
.Pp
Exit from the shell, and the jail will be shut down.
2001-12-14 10:18:15 +00:00
.Ss "Starting the Jail"
You are now ready to restart the jail and bring up the environment with
2001-12-14 10:18:15 +00:00
all of its daemons and other programs.
Create an entry for the jail in
.Pa /etc/jail.conf :
.Bd -literal -offset indent
testjail {
path = /tmp/jail/testjail;
mount.devfs;
host.hostname = testhostname;
ip4.addr = 192.0.2.100;
interface = em0;
exec.start = "/bin/sh /etc/rc";
add ability to set watchdog timeout for a shutdown This change allows to specify a watchdog(9) timeout for a system shutdown. The timeout is activated when the watchdogd daemon is stopped. The idea is to a prevent any indefinite hang during late stages of the shutdown. The feature is implemented in rc.d/watchdogd, it builds upon watchdogd -x option. Note that the shutdown timeout is not actiavted when the watchdogd service is individually stopped by an operator. It is also not activated for the 'shutdown' to the single-user mode. In those cases it is assumed that the operator knows what they are doing and they have means to recover the system should it hang. Significant subchanges and implementation details: - the argument to rc.shutdown, completely unused before, is assigned to rc_shutdown variable that can be inspected by rc scripts - init(8) passes "single" or "reboot" as the argument, this is not changed - the argument is not mandatory and if it is not set then rc_shutdown is set to "unspecified" - however, the default jail management scripts and jail configuration examples have been updated to pass "jail" to rc.shutdown, just in case - the new timeout can be set via watchdogd_shutdown_timeout rc option - for consistency, the regular timeout can now be set via watchdogd_timeout rc option - watchdogd_shutdown_timeout and watchdogd_timeout override timeout specifications in watchdogd_flags - existing configurations, where the new rc options are not set, should keep working as before I am not particularly wed to any of the implementation specifics. I am open to changing or removing any of them as long as the provided functionality is the same (or very close) to the proposed one. For example, I think it can be implemented without using watchdogd -x, by means of watchdog(1) alone. In that case there would be a small window between stopping watchdogd and running watchdog, but I think that that is acceptable. Reviewed by: bcr (man page changes) MFC after: 5 weeks Relnotes: yes Differential Revision: https://reviews.freebsd.org/D21221
2019-10-03 11:23:10 +00:00
exec.stop = "/bin/sh /etc/rc.shutdown jail";
}
.Ed
.Pp
To start a virtual server environment,
.Pa /etc/rc
is run to launch various daemons and services, and
.Pa /etc/rc.shutdown
is run to shut them down when the jail is removed.
If you are running a single application in the jail,
substitute the command used to start the application for
.Dq /bin/sh /etc/rc ;
there may be some script available to cleanly shut down the application,
or it may be sufficient to go without a stop command, and have
.Nm
send
.Dv SIGTERM
to the application.
.Pp
Start the jail by running:
.Bd -literal -offset indent
jail -c testjail
.Ed
.Pp
A few warnings may be produced; however, it should all work properly.
You should be able to see
.Xr inetd 8 ,
.Xr syslogd 8 ,
and other processes running within the jail using
.Xr ps 1 ,
with the
2001-12-14 10:18:15 +00:00
.Ql J
flag appearing beside jailed processes.
To see an active list of jails, use
.Xr jls 8 .
If
.Xr sshd 8
is enabled in the jail environment, you should be able to
.Xr ssh 1
2001-12-14 10:18:15 +00:00
to the hostname or IP address of the jailed environment, and log
in using the accounts you created previously.
.Pp
It is possible to have jails started at boot time.
Please refer to the
.Dq jail_*
variables in
.Xr rc.conf 5
for more information.
2001-12-14 10:18:15 +00:00
.Ss "Managing the Jail"
Normal machine shutdown commands, such as
.Xr halt 8 ,
.Xr reboot 8 ,
and
.Xr shutdown 8 ,
2001-12-14 10:18:15 +00:00
cannot be used successfully within the jail.
To kill all processes from within a jail, you may use one of the
following commands, depending on what you want to accomplish:
2001-12-14 10:18:15 +00:00
.Bd -literal -offset indent
kill -TERM -1
kill -KILL -1
.Ed
.Pp
This will send the
2001-12-14 10:18:15 +00:00
.Dv SIGTERM
or
2001-12-14 10:18:15 +00:00
.Dv SIGKILL
signals to all processes in the jail \(em be careful not to run this from
the host environment!
Once all of the jail's processes have died, unless the jail was created
with the
.Va persist
parameter, the jail will be removed.
2001-12-14 10:18:15 +00:00
Depending on
the intended use of the jail, you may also want to run
.Pa /etc/rc.shutdown
2001-12-14 10:18:15 +00:00
from within the jail.
.Pp
To shut down the jail from the outside, simply remove it with
.Nm
.Ar -r ,
which will run any commands specified by
.Va exec.stop ,
and then send
.Dv SIGTERM
and eventually
.Dv SIGKILL
to any remaining jailed processes.
.Pp
2001-12-14 10:18:15 +00:00
The
.Pa /proc/ Ns Ar pid Ns Pa /status
file contains, as its last field, the name of the jail in which the
process runs, or
2001-12-14 10:18:15 +00:00
.Dq Li -
to indicate that the process is not running within a jail.
The
.Xr ps 1
command also shows a
2001-12-14 10:18:15 +00:00
.Ql J
flag for processes in a jail.
.Pp
You can also list/kill processes based on their jail ID.
To show processes and their jail ID, use the following command:
.Pp
.Dl "ps ax -o pid,jid,args"
.Pp
To show and then kill processes in jail number 3 use the following commands:
.Bd -literal -offset indent
pgrep -lfj 3
pkill -j 3
.Ed
or:
.Pp
.Dl "killall -j 3"
.Ss "Jails and File Systems"
It is not possible to
.Xr mount 8
or
.Xr umount 8
any file system inside a jail unless the file system is marked
jail-friendly, the jail's
.Va allow.mount
parameter is set, and the jail's
.Va enforce_statfs
parameter is lower than 2.
.Pp
Multiple jails sharing the same file system can influence each other.
For example, a user in one jail can fill the file system,
leaving no space for processes in the other jail.
Trying to use
.Xr quota 1
to prevent this will not work either, as the file system quotas
are not aware of jails but only look at the user and group IDs.
This means the same user ID in two jails share a single file
system quota.
One would need to use one file system per jail to make this work.
2001-12-14 10:18:15 +00:00
.Ss "Sysctl MIB Entries"
The read-only entry
.Va security.jail.jailed
2006-09-29 17:57:04 +00:00
can be used to determine if a process is running inside a jail (value
is one) or not (value is zero).
.Pp
The variable
.Va security.jail.max_af_ips
determines how may address per address family a jail may have.
The default is 255.
.Pp
Some MIB variables have per-jail settings.
Changes to these variables by a jailed process do not affect the host
environment, only the jail environment.
These variables are
.Va kern.securelevel ,
.Va security.bsd.suser_enabled ,
.Va kern.hostname ,
.Va kern.domainname ,
.Va kern.hostid ,
and
.Va kern.hostuuid .
.Ss "Hierarchical Jails"
By setting a jail's
.Va children.max
parameter, processes within a jail may be able to create jails of their own.
These child jails are kept in a hierarchy, with jails only able to see and/or
modify the jails they created (or those jails' children).
Each jail has a read-only
.Va parent
parameter, containing the
.Va jid
of the jail that created it; a
.Va jid
of 0 indicates the jail is a child of the current jail (or is a top-level
jail if the current process isn't jailed).
.Pp
Jailed processes are not allowed to confer greater permissions than they
themselves are given, e.g., if a jail is created with
.Va allow.nomount ,
it is not able to create a jail with
.Va allow.mount
set.
Similarly, such restrictions as
.Va ip4.addr
and
.Va securelevel
may not be bypassed in child jails.
.Pp
A child jail may in turn create its own child jails if its own
.Va children.max
parameter is set (remember it is zero by default).
These jails are visible to and can be modified by their parent and all
ancestors.
.Pp
Jail names reflect this hierarchy, with a full name being an MIB-type string
separated by dots.
For example, if a base system process creates a jail
.Dq foo ,
and a process under that jail creates another jail
.Dq bar ,
then the second jail will be seen as
.Dq foo.bar
in the base system (though it is only seen as
.Dq bar
to any processes inside jail
.Dq foo ) .
Jids on the other hand exist in a single space, and each jail must have a
unique jid.
.Pp
Like the names, a child jail's
.Va path
appears relative to its creator's own
.Va path .
This is by virtue of the child jail being created in the chrooted
environment of the first jail.
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Sh SEE ALSO
.Xr killall 1 ,
.Xr lsvfs 1 ,
.Xr newaliases 1 ,
.Xr pgrep 1 ,
.Xr pkill 1 ,
.Xr ps 1 ,
.Xr quota 1 ,
.Xr jail_set 2 ,
.Xr vmm 4 ,
.Xr devfs 5 ,
.Xr fdescfs 5 ,
.Xr jail.conf 5 ,
.Xr linprocfs 5 ,
.Xr linsysfs 5 ,
.Xr procfs 5 ,
.Xr rc.conf 5 ,
.Xr sysctl.conf 5 ,
.Xr bsdconfig 8 ,
.Xr chroot 8 ,
.Xr devfs 8 ,
.Xr halt 8 ,
2014-12-26 21:56:23 +00:00
.Xr ifconfig 8 ,
.Xr inetd 8 ,
.Xr jexec 8 ,
.Xr jls 8 ,
.Xr mount 8 ,
.Xr named 8 ,
.Xr reboot 8 ,
2001-07-05 08:13:03 +00:00
.Xr rpcbind 8 ,
.Xr sendmail 8 ,
.Xr shutdown 8 ,
.Xr sysctl 8 ,
2009-01-12 07:45:03 +00:00
.Xr syslogd 8 ,
.Xr umount 8
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Sh HISTORY
The
.Nm
2002-07-14 14:47:15 +00:00
utility appeared in
This Implements the mumbled about "Jail" feature. This is a seriously beefed up chroot kind of thing. The process is jailed along the same lines as a chroot does it, but with additional tough restrictions imposed on what the superuser can do. For all I know, it is safe to hand over the root bit inside a prison to the customer living in that prison, this is what it was developed for in fact: "real virtual servers". Each prison has an ip number associated with it, which all IP communications will be coerced to use and each prison has its own hostname. Needless to say, you need more RAM this way, but the advantage is that each customer can run their own particular version of apache and not stomp on the toes of their neighbors. It generally does what one would expect, but setting up a jail still takes a little knowledge. A few notes: I have no scripts for setting up a jail, don't ask me for them. The IP number should be an alias on one of the interfaces. mount a /proc in each jail, it will make ps more useable. /proc/<pid>/status tells the hostname of the prison for jailed processes. Quotas are only sensible if you have a mountpoint per prison. There are no privisions for stopping resource-hogging. Some "#ifdef INET" and similar may be missing (send patches!) If somebody wants to take it from here and develop it into more of a "virtual machine" they should be most welcome! Tools, comments, patches & documentation most welcome. Have fun... Sponsored by: http://www.rndassociates.com/ Run for almost a year by: http://www.servetheweb.com/
1999-04-28 11:38:52 +00:00
.Fx 4.0 .
Hierarchical/extensible jails were introduced in
.Fx 8.0 .
The configuration file was introduced in
.Fx 9.1 .
1999-12-21 11:25:10 +00:00
.Sh AUTHORS
2001-12-14 10:18:15 +00:00
.An -nosplit
The jail feature was written by
.An Poul-Henning Kamp
for R&D Associates
2000-11-14 11:20:58 +00:00
who contributed it to
.Fx .
.Pp
2001-12-14 10:18:15 +00:00
.An Robert Watson
wrote the extended documentation, found a few bugs, added
a few new features, and cleaned up the userland jail environment.
MFp4: Bring in updated jail support from bz_jail branch. This enhances the current jail implementation to permit multiple addresses per jail. In addtion to IPv4, IPv6 is supported as well. Due to updated checks it is even possible to have jails without an IP address at all, which basically gives one a chroot with restricted process view, no networking,.. SCTP support was updated and supports IPv6 in jails as well. Cpuset support permits jails to be bound to specific processor sets after creation. Jails can have an unrestricted (no duplicate protection, etc.) name in addition to the hostname. The jail name cannot be changed from within a jail and is considered to be used for management purposes or as audit-token in the future. DDB 'show jails' command was added to aid debugging. Proper compat support permits 32bit jail binaries to be used on 64bit systems to manage jails. Also backward compatibility was preserved where possible: for jail v1 syscalls, as well as with user space management utilities. Both jail as well as prison version were updated for the new features. A gap was intentionally left as the intermediate versions had been used by various patches floating around the last years. Bump __FreeBSD_version for the afore mentioned and in kernel changes. Special thanks to: - Pawel Jakub Dawidek (pjd) for his multi-IPv4 patches and Olivier Houchard (cognet) for initial single-IPv6 patches. - Jeff Roberson (jeff) and Randall Stewart (rrs) for their help, ideas and review on cpuset and SCTP support. - Robert Watson (rwatson) for lots and lots of help, discussions, suggestions and review of most of the patch at various stages. - John Baldwin (jhb) for his help. - Simon L. Nielsen (simon) as early adopter testing changes on cluster machines as well as all the testers and people who provided feedback the last months on freebsd-jail and other channels. - My employer, CK Software GmbH, for the support so I could work on this. Reviewed by: (see above) MFC after: 3 months (this is just so that I get the mail) X-MFC Before: 7.2-RELEASE if possible
2008-11-29 14:32:14 +00:00
.Pp
.An Bjoern A. Zeeb
added multi-IP jail support for IPv4 and IPv6 based on a patch
originally done by
.An Pawel Jakub Dawidek
for IPv4.
.Pp
.An James Gritton
added the extensible jail parameters, hierarchical jails,
and the configuration file.
.Sh BUGS
It might be a good idea to add an
2001-12-14 10:18:15 +00:00
address alias flag such that daemons listening on all IPs
.Pq Dv INADDR_ANY
will not bind on that address, which would facilitate building a safe
host environment such that host daemons do not impose on services offered
2001-12-14 10:18:15 +00:00
from within jails.
2004-05-20 06:37:44 +00:00
Currently, the simplest answer is to minimize services
offered on the host, possibly limiting it to services offered from
.Xr inetd 8
which is easily configurable.
.Sh NOTES
Great care should be taken when managing directories visible within the jail.
For example, if a jailed process has its current working directory set to a
directory that is moved out of the jail's chroot, then the process may gain
access to the file space outside of the jail.
It is recommended that directories always be copied, rather than moved, out
of a jail.
2012-09-16 15:22:15 +00:00
.Pp
In addition, there are several ways in which an unprivileged user
outside the jail can cooperate with a privileged user inside the jail
and thereby obtain elevated privileges in the host environment.
Most of these attacks can be mitigated by ensuring that the jail root
is not accessible to unprivileged users in the host environment.
Regardless, as a general rule, untrusted users with privileged access
to a jail should not be given access to the host environment.