4 include::attributes.txt[]
9 pct - Tool to manage Linux Containers (LXC) on Proxmox VE
15 include::pct.1-synopsis.adoc[]
22 Proxmox Container Toolkit
23 =========================
24 include::attributes.txt[]
28 Containers are a lightweight alternative to fully virtualized
29 VMs. Instead of emulating a complete Operating System (OS), containers
30 simply use the OS of the host they run on. This implies that all
31 containers use the same kernel, and that they can access resources
32 from the host directly.
34 This is great because containers do not waste CPU power nor memory due
35 to kernel emulation. Container run-time costs are close to zero and
36 usually negligible. But there are also some drawbacks you need to
39 * You can only run Linux based OS inside containers, i.e. it is not
40 possible to run FreeBSD or MS Windows inside.
42 * For security reasons, access to host resources needs to be
43 restricted. This is done with AppArmor, SecComp filters and other
44 kernel features. Be prepared that some syscalls are not allowed
47 {pve} uses https://linuxcontainers.org/[LXC] as underlying container
48 technology. We consider LXC as low-level library, which provides
49 countless options. It would be too difficult to use those tools
50 directly. Instead, we provide a small wrapper called `pct`, the
51 "Proxmox Container Toolkit".
53 The toolkit is tightly coupled with {pve}. That means that it is aware
54 of the cluster setup, and it can use the same network and storage
55 resources as fully virtualized VMs. You can even use the {pve}
56 firewall, or manage containers using the HA framework.
58 Our primary goal is to offer an environment as one would get from a
59 VM, but without the additional overhead. We call this "System
62 NOTE: If you want to run micro-containers (with docker, rkt, ...), it
63 is best to run them inside a VM.
66 Security Considerations
67 -----------------------
69 Containers use the same kernel as the host, so there is a big attack
70 surface for malicious users. You should consider this fact if you
71 provide containers to totally untrusted people. In general, fully
72 virtualized VMs provide better isolation.
74 The good news is that LXC uses many kernel security features like
75 AppArmor, CGroups and PID and user namespaces, which makes containers
76 usage quite secure. We distinguish two types of containers:
81 Security is done by dropping capabilities, using mandatory access
82 control (AppArmor), SecComp filters and namespaces. The LXC team
83 considers this kind of container as unsafe, and they will not consider
84 new container escape exploits to be security issues worthy of a CVE
85 and quick fix. So you should use this kind of containers only inside a
86 trusted environment, or when no untrusted task is running as root in
89 Unprivileged containers
90 ~~~~~~~~~~~~~~~~~~~~~~~
92 This kind of containers use a new kernel feature called user
93 namespaces. The root uid 0 inside the container is mapped to an
94 unprivileged user outside the container. This means that most security
95 issues (container escape, resource abuse, ...) in those containers
96 will affect a random unprivileged user, and so would be a generic
97 kernel security bug rather than an LXC issue. The LXC team thinks
98 unprivileged containers are safe by design.
104 The '/etc/pve/lxc/<CTID>.conf' file stores container configuration,
105 where '<CTID>' is the numeric ID of the given container. Like all
106 other files stored inside '/etc/pve/', they get automatically
107 replicated to all other cluster nodes.
109 NOTE: CTIDs < 100 are reserved for internal purposes, and CTIDs need to be
112 .Example Container Configuration
119 net0: bridge=vmbr0,hwaddr=66:64:66:64:64:36,ip=dhcp,name=eth0,type=veth
120 rootfs: local:107/vm-107-disk-1.raw,size=7G
123 Those configuration files are simple text files, and you can edit them
124 using a normal text editor ('vi', 'nano', ...). This is sometimes
125 useful to do small corrections, but keep in mind that you need to
126 restart the container to apply such changes.
128 For that reason, it is usually better to use the 'pct' command to
129 generate and modify those files, or do the whole thing using the GUI.
130 Our toolkit is smart enough to instantaneously apply most changes to
131 running containers. This feature is called "hot plug", and there is no
132 need to restart the container in that case.
137 Container configuration files use a simple colon separated key/value
138 format. Each line has the following format:
143 Blank lines in those files are ignored, and lines starting with a '#'
144 character are treated as comments and are also ignored.
146 It is possible to add low-level, LXC style configuration directly, for
149 lxc.init_cmd: /sbin/my_own_init
153 lxc.init_cmd = /sbin/my_own_init
155 Those settings are directly passed to the LXC low-level tools.
160 When you create a snapshot, 'pct' stores the configuration at snapshot
161 time into a separate snapshot section within the same configuration
162 file. For example, after creating a snapshot called 'testsnapshot',
163 your configuration file will look like this:
165 .Container Configuration with Snapshot
179 There are a few snapshot related properties like 'parent' and
180 'snaptime'. The 'parent' property is used to store the parent/child
181 relationship between snapshots. 'snaptime' is the snapshot creation
182 time stamp (unix epoch).
184 Guest Operating System Configuration
185 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
187 We normally try to detect the operating system type inside the
188 container, and then modify some files inside the container to make
189 them work as expected. Here is a short list of things we do at
192 set /etc/hostname:: to set the container name
194 modify /etc/hosts:: to allow lookup of the local hostname
196 network setup:: pass the complete network setup to the container
198 configure DNS:: pass information about DNS servers
200 adapt the init system:: for example, fix the number of spawned getty processes
202 set the root password:: when creating a new container
204 rewrite ssh_host_keys:: so that each container has unique keys
206 randomize crontab:: so that cron does not start at the same time on all containers
208 Changes made by {PVE} are enclosed by comment markers:
216 Those markers will be inserted at a reasonable location in the
217 file. If such a section already exists, it will be updated in place
218 and will not be moved.
220 Modification of a file can be prevented by adding a `.pve-ignore.`
221 file for it. For instance, if the file `/etc/.pve-ignore.hosts`
222 exists then the `/etc/hosts` file will not be touched. This can be a
223 simple empty file creatd via:
225 # touch /etc/.pve-ignore.hosts
227 Most modifications are OS dependent, so they differ between different
228 distributions and versions. You can completely disable modifications
229 by manually setting the 'ostype' to 'unmanaged'.
231 OS type detection is done by testing for certain files inside the
234 Ubuntu:: inspect /etc/lsb-release ('DISTRIB_ID=Ubuntu')
236 Debian:: test /etc/debian_version
238 Fedora:: test /etc/fedora-release
240 RedHat or CentOS:: test /etc/redhat-release
242 ArchLinux:: test /etc/arch-release
244 Alpine:: test /etc/alpine-release
246 Gentoo:: test /etc/gentoo-release
248 NOTE: Container start fails if the configured 'ostype' differs from the auto
254 include::pct.conf.5-opts.adoc[]
260 Container Images, sometimes also referred to as "templates" or
261 "appliances", are 'tar' archives which contain everything to run a
262 container. You can think of it as a tidy container backup. Like most
263 modern container toolkits, 'pct' uses those images when you create a
264 new container, for example:
266 pct create 999 local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz
268 Proxmox itself ships a set of basic templates for most common
269 operating systems, and you can download them using the 'pveam' (short
270 for {pve} Appliance Manager) command line utility. You can also
271 download https://www.turnkeylinux.org/[TurnKey Linux] containers using
272 that tool (or the graphical user interface).
274 Our image repositories contain a list of available images, and there
275 is a cron job run each day to download that list. You can trigger that
276 update manually with:
280 After that you can view the list of available images using:
284 You can restrict this large list by specifying the 'section' you are
285 interested in, for example basic 'system' images:
287 .List available system images
289 # pveam available --section system
290 system archlinux-base_2015-24-29-1_x86_64.tar.gz
291 system centos-7-default_20160205_amd64.tar.xz
292 system debian-6.0-standard_6.0-7_amd64.tar.gz
293 system debian-7.0-standard_7.0-3_amd64.tar.gz
294 system debian-8.0-standard_8.0-1_amd64.tar.gz
295 system ubuntu-12.04-standard_12.04-1_amd64.tar.gz
296 system ubuntu-14.04-standard_14.04-1_amd64.tar.gz
297 system ubuntu-15.04-standard_15.04-1_amd64.tar.gz
298 system ubuntu-15.10-standard_15.10-1_amd64.tar.gz
301 Before you can use such a template, you need to download them into one
302 of your storages. You can simply use storage 'local' for that
303 purpose. For clustered installations, it is preferred to use a shared
304 storage so that all nodes can access those images.
306 pveam download local debian-8.0-standard_8.0-1_amd64.tar.gz
308 You are now ready to create containers using that image, and you can
309 list all downloaded images on storage 'local' with:
313 local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz 190.20MB
316 The above command shows you the full {pve} volume identifiers. They include
317 the storage name, and most other {pve} commands can use them. For
318 examply you can delete that image later with:
320 pveam remove local:vztmpl/debian-8.0-standard_8.0-1_amd64.tar.gz
326 Traditional containers use a very simple storage model, only allowing
327 a single mount point, the root file system. This was further
328 restricted to specific file system types like 'ext4' and 'nfs'.
329 Additional mounts are often done by user provided scripts. This turend
330 out to be complex and error prone, so we try to avoid that now.
332 Our new LXC based container model is more flexible regarding
333 storage. First, you can have more than a single mount point. This
334 allows you to choose a suitable storage for each application. For
335 example, you can use a relatively slow (and thus cheap) storage for
336 the container root file system. Then you can use a second mount point
337 to mount a very fast, distributed storage for your database
340 The second big improvement is that you can use any storage type
341 supported by the {pve} storage library. That means that you can store
342 your containers on local 'lvmthin' or 'zfs', shared 'iSCSI' storage,
343 or even on distributed storage systems like 'ceph'. It also enables us
344 to use advanced storage features like snapshots and clones. 'vzdump'
345 can also use the snapshot feature to provide consistent container
348 Last but not least, you can also mount local devices directly, or
349 mount local directories using bind mounts. That way you can access
350 local storage inside containers with zero overhead. Such bind mounts
351 also provide an easy way to share data between different containers.
357 The root mount point is configured with the `rootfs` property, and you can
358 configure up to 10 additional mount points. The corresponding options
359 are called `mp0` to `mp9`, and they can contain the following setting:
361 include::pct-mountpoint-opts.adoc[]
363 Currently there are basically three types of mount points: storage backed
364 mount points, bind mounts and device mounts.
366 .Typical Container `rootfs` configuration
368 rootfs: thin1:base-100-disk-1,size=8G
372 Storage backed mount points
373 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
375 Storage backed mount points are managed by the {pve} storage subsystem and come
376 in three different flavors:
378 - Image based: These are raw images containing a single ext4 formatted file
380 - ZFS Subvolumes: These are technically bind mounts, but with managed storage,
381 and thus allow resizing and snapshotting.
382 - Directories: passing `size=0` triggers a special case where instead of a raw
383 image a directory is created.
389 Bind mounts are considered to not be managed by the storage subsystem, so you
390 cannot make snapshots or deal with quotas from inside the container, and with
391 unprivileged containers you might run into permission problems caused by the
392 user mapping, and cannot use ACLs from inside an unprivileged container.
394 WARNING: For security reasons, bind mounts should only be established
395 using source directories especially reserved for this purpose, e.g., a
396 directory hierarchy under `/mnt/bindmounts`. Never bind mount system
397 directories like `/`, `/var` or `/etc` into a container - this poses a
398 great security risk. The bind mount source path must not contain any symlinks.
404 Similar to bind mounts, device mounts are not managed by the storage, but for
405 these the `quota` and `acl` options will be honored.
411 WARNING: Because of existing issues in the Linux kernel's freezer
412 subsystem the usage of FUSE mounts inside a container is strongly
413 advised against, as containers need to be frozen for suspend or
414 snapshot mode backups.
416 If FUSE mounts cannot be replaced by other mounting mechanisms or storage
417 technologies, it is possible to establish the FUSE mount on the Proxmox host
418 and use a bind mount point to make it accessible inside the container.
421 Using quotas inside containers
422 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
424 Quotas allow to set limits inside a container for the amount of disk
425 space that each user can use. This only works on ext4 image based
426 storage types and currently does not work with unprivileged
429 Activating the `quota` option causes the following mount options to be
430 used for a mount point:
431 `usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0`
433 This allows quotas to be used like you would on any other system. You
434 can initialize the `/aquota.user` and `/aquota.group` files by running
441 and edit the quotas via the `edquota` command. Refer to the documentation
442 of the distribution running inside the container for details.
444 NOTE: You need to run the above commands for every mount point by passing
445 the mount point's path instead of just `/`.
448 Using ACLs inside containers
449 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
451 The standard Posix Access Control Lists are also available inside containers.
452 ACLs allow you to set more detailed file ownership than the traditional user/
459 You can configure up to 10 network interfaces for a single
460 container. The corresponding options are called 'net0' to 'net9', and
461 they can contain the following setting:
463 include::pct-network-opts.adoc[]
472 It is possible to use the 'vzdump' tool for container backup. Please
473 refer to the 'vzdump' manual page for details.
475 Restoring Container Backups
476 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
478 Restoring container backups made with 'vzdump' is possible using the
479 'pct restore' command. By default, 'pct restore' will attempt to restore as much
480 of the backed up container configuration as possible. It is possible to override
481 the backed up configuration by manually setting container options on the command
482 line (see the 'pct' manual page for details).
484 NOTE: 'pvesm extractconfig' can be used to view the backed up configuration
485 contained in a vzdump archive.
487 There are two basic restore modes, only differing by their handling of mount
491 "Simple" restore mode
492 ^^^^^^^^^^^^^^^^^^^^^
494 If neither the `rootfs` parameter nor any of the optional `mpX` parameters
495 are explicitly set, the mount point configuration from the backed up
496 configuration file is restored using the following steps:
498 . Extract mount points and their options from backup
499 . Create volumes for storage backed mount points (on storage provided with the
500 `storage` parameter, or default local storage if unset)
501 . Extract files from backup archive
502 . Add bind and device mount points to restored configuration (limited to root user)
504 NOTE: Since bind and device mount points are never backed up, no files are
505 restored in the last step, but only the configuration options. The assumption
506 is that such mount points are either backed up with another mechanism (e.g.,
507 NFS space that is bind mounted into many containers), or not intended to be
510 This simple mode is also used by the container restore operations in the web
514 "Advanced" restore mode
515 ^^^^^^^^^^^^^^^^^^^^^^^
517 By setting the `rootfs` parameter (and optionally, any combination of `mpX`
518 parameters), the 'pct restore' command is automatically switched into an
519 advanced mode. This advanced mode completely ignores the `rootfs` and `mpX`
520 configuration options contained in the backup archive, and instead only
521 uses the options explicitly provided as parameters.
523 This mode allows flexible configuration of mount point settings at restore time,
526 * Set target storages, volume sizes and other options for each mount point
528 * Redistribute backed up files according to new mount point scheme
529 * Restore to device and/or bind mount points (limited to root user)
532 Managing Containers with 'pct'
533 ------------------------------
535 'pct' is the tool to manage Linux Containers on {pve}. You can create
536 and destroy containers, and control execution (start, stop, migrate,
537 ...). You can use pct to set parameters in the associated config file,
538 like network configuration or memory limits.
543 Create a container based on a Debian template (provided you have
544 already downloaded the template via the webgui)
546 pct create 100 /var/lib/vz/template/cache/debian-8.0-standard_8.0-1_amd64.tar.gz
552 Start a login session via getty
556 Enter the LXC namespace and run a shell as root user
560 Display the configuration
564 Add a network interface called eth0, bridged to the host bridge vmbr0,
565 set the address and gateway, while it's running
567 pct set 100 -net0 name=eth0,bridge=vmbr0,ip=192.168.15.147/24,gw=192.168.15.1
569 Reduce the memory of the container to 512MB
571 pct set 100 -memory 512
577 '/etc/pve/lxc/<CTID>.conf'::
579 Configuration file for the container '<CTID>'.
585 - Simple, and fully integrated into {pve}. Setup looks similar to a normal
588 * Storage (ZFS, LVM, NFS, Ceph, ...)
596 - Fast: minimal overhead, as fast as bare metal
598 - High density (perfect for idle workloads)
602 - Direct hardware access
608 - Integrated into {pve} graphical user interface (GUI)
610 - LXC (https://linuxcontainers.org/)
612 - cgmanager for cgroup management
614 - lxcfs to provive containerized /proc file system
618 - CRIU: for live migration (planned)
620 - We use latest available kernels (4.4.X)
622 - Image based deployment (templates)
624 - Container setup from host (Network, DNS, Storage, ...)
628 include::pve-copyright.adoc[]