5 include::attributes.txt[]
11 pct - Tool to manage Linux Containers (LXC) on Proxmox VE
17 include::pct.1-synopsis.adoc[]
24 Proxmox Container Toolkit
25 =========================
26 include::attributes.txt[]
30 :title: Linux Container
33 Containers are a lightweight alternative to fully virtualized
34 VMs. Instead of emulating a complete Operating System (OS), containers
35 simply use the OS of the host they run on. This implies that all
36 containers use the same kernel, and that they can access resources
37 from the host directly.
39 This is great because containers do not waste CPU power nor memory due
40 to kernel emulation. Container run-time costs are close to zero and
41 usually negligible. But there are also some drawbacks you need to
44 * You can only run Linux based OS inside containers, i.e. it is not
45 possible to run FreeBSD or MS Windows inside.
47 * For security reasons, access to host resources needs to be
48 restricted. This is done with AppArmor, SecComp filters and other
49 kernel features. Be prepared that some syscalls are not allowed
52 {pve} uses https://linuxcontainers.org/[LXC] as underlying container
53 technology. We consider LXC as low-level library, which provides
54 countless options. It would be too difficult to use those tools
55 directly. Instead, we provide a small wrapper called `pct`, the
56 "Proxmox Container Toolkit".
58 The toolkit is tightly coupled with {pve}. That means that it is aware
59 of the cluster setup, and it can use the same network and storage
60 resources as fully virtualized VMs. You can even use the {pve}
61 firewall, or manage containers using the HA framework.
63 Our primary goal is to offer an environment as one would get from a
64 VM, but without the additional overhead. We call this "System
67 NOTE: If you want to run micro-containers (with docker, rkt, ...), it
68 is best to run them inside a VM.
74 * LXC (https://linuxcontainers.org/)
76 * Integrated into {pve} graphical user interface (GUI)
78 * Easy to use command line tool `pct`
80 * Access via {pve} REST API
82 * lxcfs to provide containerized /proc file system
84 * AppArmor/Seccomp to improve security
86 * CRIU: for live migration (planned)
88 * Use latest available kernels (4.4.X)
90 * Image based deployment (templates)
92 * Use {pve} storage library
94 * Container setup from host (network, DNS, storage, ...)
97 Security Considerations
98 -----------------------
100 Containers use the same kernel as the host, so there is a big attack
101 surface for malicious users. You should consider this fact if you
102 provide containers to totally untrusted people. In general, fully
103 virtualized VMs provide better isolation.
105 The good news is that LXC uses many kernel security features like
106 AppArmor, CGroups and PID and user namespaces, which makes containers
107 usage quite secure. We distinguish two types of containers:
110 Privileged Containers
111 ~~~~~~~~~~~~~~~~~~~~~
113 Security is done by dropping capabilities, using mandatory access
114 control (AppArmor), SecComp filters and namespaces. The LXC team
115 considers this kind of container as unsafe, and they will not consider
116 new container escape exploits to be security issues worthy of a CVE
117 and quick fix. So you should use this kind of containers only inside a
118 trusted environment, or when no untrusted task is running as root in
122 Unprivileged Containers
123 ~~~~~~~~~~~~~~~~~~~~~~~
125 This kind of containers use a new kernel feature called user
126 namespaces. The root UID 0 inside the container is mapped to an
127 unprivileged user outside the container. This means that most security
128 issues (container escape, resource abuse, ...) in those containers
129 will affect a random unprivileged user, and so would be a generic
130 kernel security bug rather than an LXC issue. The LXC team thinks
131 unprivileged containers are safe by design.
134 Guest Operating System Configuration
135 ------------------------------------
137 We normally try to detect the operating system type inside the
138 container, and then modify some files inside the container to make
139 them work as expected. Here is a short list of things we do at
142 set /etc/hostname:: to set the container name
144 modify /etc/hosts:: to allow lookup of the local hostname
146 network setup:: pass the complete network setup to the container
148 configure DNS:: pass information about DNS servers
150 adapt the init system:: for example, fix the number of spawned getty processes
152 set the root password:: when creating a new container
154 rewrite ssh_host_keys:: so that each container has unique keys
156 randomize crontab:: so that cron does not start at the same time on all containers
158 Changes made by {PVE} are enclosed by comment markers:
166 Those markers will be inserted at a reasonable location in the
167 file. If such a section already exists, it will be updated in place
168 and will not be moved.
170 Modification of a file can be prevented by adding a `.pve-ignore.`
171 file for it. For instance, if the file `/etc/.pve-ignore.hosts`
172 exists then the `/etc/hosts` file will not be touched. This can be a
173 simple empty file creatd via:
175 # touch /etc/.pve-ignore.hosts
177 Most modifications are OS dependent, so they differ between different
178 distributions and versions. You can completely disable modifications
179 by manually setting the `ostype` to `unmanaged`.
181 OS type detection is done by testing for certain files inside the
184 Ubuntu:: inspect /etc/lsb-release (`DISTRIB_ID=Ubuntu`)
186 Debian:: test /etc/debian_version
188 Fedora:: test /etc/fedora-release
190 RedHat or CentOS:: test /etc/redhat-release
192 ArchLinux:: test /etc/arch-release
194 Alpine:: test /etc/alpine-release
196 Gentoo:: test /etc/gentoo-release
198 NOTE: Container start fails if the configured `ostype` differs from the auto
202 [[pct_container_images]]
206 Container images, sometimes also referred to as ``templates'' or
207 ``appliances'', are `tar` archives which contain everything to run a
208 container. You can think of it as a tidy container backup. Like most
209 modern container toolkits, `pct` uses those images when you create a
210 new container, for example:
212 pct create 999 local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz
214 {pve} itself ships a set of basic templates for most common
215 operating systems, and you can download them using the `pveam` (short
216 for {pve} Appliance Manager) command line utility. You can also
217 download https://www.turnkeylinux.org/[TurnKey Linux] containers using
218 that tool (or the graphical user interface).
220 Our image repositories contain a list of available images, and there
221 is a cron job run each day to download that list. You can trigger that
222 update manually with:
226 After that you can view the list of available images using:
230 You can restrict this large list by specifying the `section` you are
231 interested in, for example basic `system` images:
233 .List available system images
235 # pveam available --section system
236 system archlinux-base_2015-24-29-1_x86_64.tar.gz
237 system centos-7-default_20160205_amd64.tar.xz
238 system debian-6.0-standard_6.0-7_amd64.tar.gz
239 system debian-7.0-standard_7.0-3_amd64.tar.gz
240 system debian-8.0-standard_8.0-1_amd64.tar.gz
241 system ubuntu-12.04-standard_12.04-1_amd64.tar.gz
242 system ubuntu-14.04-standard_14.04-1_amd64.tar.gz
243 system ubuntu-15.04-standard_15.04-1_amd64.tar.gz
244 system ubuntu-15.10-standard_15.10-1_amd64.tar.gz
247 Before you can use such a template, you need to download them into one
248 of your storages. You can simply use storage `local` for that
249 purpose. For clustered installations, it is preferred to use a shared
250 storage so that all nodes can access those images.
252 pveam download local debian-8.0-standard_8.0-1_amd64.tar.gz
254 You are now ready to create containers using that image, and you can
255 list all downloaded images on storage `local` with:
259 local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz 190.20MB
262 The above command shows you the full {pve} volume identifiers. They include
263 the storage name, and most other {pve} commands can use them. For
264 example you can delete that image later with:
266 pveam remove local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz
269 [[pct_container_storage]]
273 Traditional containers use a very simple storage model, only allowing
274 a single mount point, the root file system. This was further
275 restricted to specific file system types like `ext4` and `nfs`.
276 Additional mounts are often done by user provided scripts. This turned
277 out to be complex and error prone, so we try to avoid that now.
279 Our new LXC based container model is more flexible regarding
280 storage. First, you can have more than a single mount point. This
281 allows you to choose a suitable storage for each application. For
282 example, you can use a relatively slow (and thus cheap) storage for
283 the container root file system. Then you can use a second mount point
284 to mount a very fast, distributed storage for your database
287 The second big improvement is that you can use any storage type
288 supported by the {pve} storage library. That means that you can store
289 your containers on local `lvmthin` or `zfs`, shared `iSCSI` storage,
290 or even on distributed storage systems like `ceph`. It also enables us
291 to use advanced storage features like snapshots and clones. `vzdump`
292 can also use the snapshot feature to provide consistent container
295 Last but not least, you can also mount local devices directly, or
296 mount local directories using bind mounts. That way you can access
297 local storage inside containers with zero overhead. Such bind mounts
298 also provide an easy way to share data between different containers.
304 The root mount point is configured with the `rootfs` property, and you can
305 configure up to 10 additional mount points. The corresponding options
306 are called `mp0` to `mp9`, and they can contain the following setting:
308 include::pct-mountpoint-opts.adoc[]
310 Currently there are basically three types of mount points: storage backed
311 mount points, bind mounts and device mounts.
313 .Typical container `rootfs` configuration
315 rootfs: thin1:base-100-disk-1,size=8G
319 Storage Backed Mount Points
320 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
322 Storage backed mount points are managed by the {pve} storage subsystem and come
323 in three different flavors:
325 - Image based: these are raw images containing a single ext4 formatted file
327 - ZFS subvolumes: these are technically bind mounts, but with managed storage,
328 and thus allow resizing and snapshotting.
329 - Directories: passing `size=0` triggers a special case where instead of a raw
330 image a directory is created.
336 Bind mounts allow you to access arbitrary directories from your Proxmox VE host
337 inside a container. Some potential use cases are:
339 - Accessing your home directory in the guest
340 - Accessing an USB device directory in the guest
341 - Accessing an NFS mount from the host in the guest
343 Bind mounts are considered to not be managed by the storage subsystem, so you
344 cannot make snapshots or deal with quotas from inside the container. With
345 unprivileged containers you might run into permission problems caused by the
346 user mapping and cannot use ACLs.
348 NOTE: The contents of bind mount points are not backed up when using `vzdump`.
350 WARNING: For security reasons, bind mounts should only be established
351 using source directories especially reserved for this purpose, e.g., a
352 directory hierarchy under `/mnt/bindmounts`. Never bind mount system
353 directories like `/`, `/var` or `/etc` into a container - this poses a
356 NOTE: The bind mount source path must not contain any symlinks.
358 For example, to make the directory `/mnt/bindmounts/shared` accessible in the
359 container with ID `100` under the path `/shared`, use a configuration line like
360 `mp0: /mnt/bindmounts/shared,mp=/shared` in `/etc/pve/lxc/100.conf`.
361 Alternatively, use `pct set 100 -mp0 /mnt/bindmounts/shared,mp=/shared` to
362 achieve the same result.
368 Device mount points allow to mount block devices of the host directly into the
369 container. Similar to bind mounts, device mounts are not managed by {PVE}'s
370 storage subsystem, but the `quota` and `acl` options will be honored.
372 NOTE: Device mount points should only be used under special circumstances. In
373 most cases a storage backed mount point offers the same performance and a lot
376 NOTE: The contents of device mount points are not backed up when using `vzdump`.
382 WARNING: Because of existing issues in the Linux kernel's freezer
383 subsystem the usage of FUSE mounts inside a container is strongly
384 advised against, as containers need to be frozen for suspend or
385 snapshot mode backups.
387 If FUSE mounts cannot be replaced by other mounting mechanisms or storage
388 technologies, it is possible to establish the FUSE mount on the Proxmox host
389 and use a bind mount point to make it accessible inside the container.
392 Using Quotas Inside Containers
393 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
395 Quotas allow to set limits inside a container for the amount of disk
396 space that each user can use. This only works on ext4 image based
397 storage types and currently does not work with unprivileged
400 Activating the `quota` option causes the following mount options to be
401 used for a mount point:
402 `usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0`
404 This allows quotas to be used like you would on any other system. You
405 can initialize the `/aquota.user` and `/aquota.group` files by running
412 and edit the quotas via the `edquota` command. Refer to the documentation
413 of the distribution running inside the container for details.
415 NOTE: You need to run the above commands for every mount point by passing
416 the mount point's path instead of just `/`.
419 Using ACLs Inside Containers
420 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
422 The standard Posix **A**ccess **C**ontrol **L**ists are also available inside containers.
423 ACLs allow you to set more detailed file ownership than the traditional user/
427 [[pct_container_network]]
431 You can configure up to 10 network interfaces for a single
432 container. The corresponding options are called `net0` to `net9`, and
433 they can contain the following setting:
435 include::pct-network-opts.adoc[]
445 It is possible to use the `vzdump` tool for container backup. Please
446 refer to the `vzdump` manual page for details.
449 Restoring Container Backups
450 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
452 Restoring container backups made with `vzdump` is possible using the
453 `pct restore` command. By default, `pct restore` will attempt to restore as much
454 of the backed up container configuration as possible. It is possible to override
455 the backed up configuration by manually setting container options on the command
456 line (see the `pct` manual page for details).
458 NOTE: `pvesm extractconfig` can be used to view the backed up configuration
459 contained in a vzdump archive.
461 There are two basic restore modes, only differing by their handling of mount
465 ``Simple'' Restore Mode
466 ^^^^^^^^^^^^^^^^^^^^^^^
468 If neither the `rootfs` parameter nor any of the optional `mpX` parameters
469 are explicitly set, the mount point configuration from the backed up
470 configuration file is restored using the following steps:
472 . Extract mount points and their options from backup
473 . Create volumes for storage backed mount points (on storage provided with the
474 `storage` parameter, or default local storage if unset)
475 . Extract files from backup archive
476 . Add bind and device mount points to restored configuration (limited to root user)
478 NOTE: Since bind and device mount points are never backed up, no files are
479 restored in the last step, but only the configuration options. The assumption
480 is that such mount points are either backed up with another mechanism (e.g.,
481 NFS space that is bind mounted into many containers), or not intended to be
484 This simple mode is also used by the container restore operations in the web
488 ``Advanced'' Restore Mode
489 ^^^^^^^^^^^^^^^^^^^^^^^^^
491 By setting the `rootfs` parameter (and optionally, any combination of `mpX`
492 parameters), the `pct restore` command is automatically switched into an
493 advanced mode. This advanced mode completely ignores the `rootfs` and `mpX`
494 configuration options contained in the backup archive, and instead only
495 uses the options explicitly provided as parameters.
497 This mode allows flexible configuration of mount point settings at restore time,
500 * Set target storages, volume sizes and other options for each mount point
502 * Redistribute backed up files according to new mount point scheme
503 * Restore to device and/or bind mount points (limited to root user)
506 Managing Containers with `pct`
507 ------------------------------
509 `pct` is the tool to manage Linux Containers on {pve}. You can create
510 and destroy containers, and control execution (start, stop, migrate,
511 ...). You can use pct to set parameters in the associated config file,
512 like network configuration or memory limits.
518 Create a container based on a Debian template (provided you have
519 already downloaded the template via the web interface)
521 pct create 100 /var/lib/vz/template/cache/debian-8.0-standard_8.0-1_amd64.tar.gz
527 Start a login session via getty
531 Enter the LXC namespace and run a shell as root user
535 Display the configuration
539 Add a network interface called `eth0`, bridged to the host bridge `vmbr0`,
540 set the address and gateway, while it's running
542 pct set 100 -net0 name=eth0,bridge=vmbr0,ip=192.168.15.147/24,gw=192.168.15.1
544 Reduce the memory of the container to 512MB
546 pct set 100 -memory 512
549 Obtaining Debugging Logs
550 ~~~~~~~~~~~~~~~~~~~~~~~~
552 In case `pct start` is unable to start a specific container, it might be
553 helpful to collect debugging output by running `lxc-start` (replace `ID` with
556 lxc-start -n ID -F -l DEBUG -o /tmp/lxc-ID.log
558 This command will attempt to start the container in foreground mode, to stop the container run `pct shutdown ID` or `pct stop ID` in a second terminal.
560 The collected debug log is written to `/tmp/lxc-ID.log`.
562 NOTE: If you have changed the container's configuration since the last start
563 attempt with `pct start`, you need to run `pct start` at least once to also
564 update the configuration used by `lxc-start`.
567 [[pct_configuration]]
571 The `/etc/pve/lxc/<CTID>.conf` file stores container configuration,
572 where `<CTID>` is the numeric ID of the given container. Like all
573 other files stored inside `/etc/pve/`, they get automatically
574 replicated to all other cluster nodes.
576 NOTE: CTIDs < 100 are reserved for internal purposes, and CTIDs need to be
579 .Example Container Configuration
586 net0: bridge=vmbr0,hwaddr=66:64:66:64:64:36,ip=dhcp,name=eth0,type=veth
587 rootfs: local:107/vm-107-disk-1.raw,size=7G
590 Those configuration files are simple text files, and you can edit them
591 using a normal text editor (`vi`, `nano`, ...). This is sometimes
592 useful to do small corrections, but keep in mind that you need to
593 restart the container to apply such changes.
595 For that reason, it is usually better to use the `pct` command to
596 generate and modify those files, or do the whole thing using the GUI.
597 Our toolkit is smart enough to instantaneously apply most changes to
598 running containers. This feature is called "hot plug", and there is no
599 need to restart the container in that case.
605 Container configuration files use a simple colon separated key/value
606 format. Each line has the following format:
613 Blank lines in those files are ignored, and lines starting with a `#`
614 character are treated as comments and are also ignored.
616 It is possible to add low-level, LXC style configuration directly, for
619 lxc.init_cmd: /sbin/my_own_init
623 lxc.init_cmd = /sbin/my_own_init
625 Those settings are directly passed to the LXC low-level tools.
632 When you create a snapshot, `pct` stores the configuration at snapshot
633 time into a separate snapshot section within the same configuration
634 file. For example, after creating a snapshot called ``testsnapshot'',
635 your configuration file will look like this:
637 .Container configuration with snapshot
651 There are a few snapshot related properties like `parent` and
652 `snaptime`. The `parent` property is used to store the parent/child
653 relationship between snapshots. `snaptime` is the snapshot creation
654 time stamp (Unix epoch).
661 include::pct.conf.5-opts.adoc[]
667 Container migrations, snapshots and backups (`vzdump`) set a lock to
668 prevent incompatible concurrent actions on the affected container. Sometimes
669 you need to remove such a lock manually (e.g., after a power failure).
673 CAUTION: Only do that if you are sure the action which set the lock is
682 `/etc/pve/lxc/<CTID>.conf`::
684 Configuration file for the container '<CTID>'.
687 include::pve-copyright.adoc[]