]>
Commit | Line | Data |
---|---|---|
80c0adcb | 1 | [[chapter_pct]] |
0c6b782f | 2 | ifdef::manvolnum[] |
b2f242ab | 3 | pct(1) |
7e2fdb3d | 4 | ====== |
5f09af76 DM |
5 | :pve-toplevel: |
6 | ||
0c6b782f DM |
7 | NAME |
8 | ---- | |
9 | ||
10 | pct - Tool to manage Linux Containers (LXC) on Proxmox VE | |
11 | ||
12 | ||
49a5e11c | 13 | SYNOPSIS |
0c6b782f DM |
14 | -------- |
15 | ||
16 | include::pct.1-synopsis.adoc[] | |
17 | ||
18 | DESCRIPTION | |
19 | ----------- | |
20 | endif::manvolnum[] | |
21 | ||
22 | ifndef::manvolnum[] | |
23 | Proxmox Container Toolkit | |
24 | ========================= | |
194d2f29 | 25 | :pve-toplevel: |
0c6b782f | 26 | endif::manvolnum[] |
5f09af76 | 27 | ifdef::wiki[] |
cb84ed18 | 28 | :title: Linux Container |
5f09af76 | 29 | endif::wiki[] |
4a2ae9ed | 30 | |
14e97811 OB |
31 | Containers are a lightweight alternative to fully virtualized machines (VMs). |
32 | They use the kernel of the host system that they run on, instead of emulating a | |
33 | full operating system (OS). This means that containers can access resources on | |
34 | the host system directly. | |
4a2ae9ed | 35 | |
6d718b9b TL |
36 | The runtime costs for containers is low, usually negligible. However, there are |
37 | some drawbacks that need be considered: | |
4a2ae9ed | 38 | |
34694a05 | 39 | * Only Linux distributions can be run in containers. It is not possible to run |
6d718b9b TL |
40 | other Operating Systems like, for example, FreeBSD or Microsoft Windows |
41 | inside a container. | |
4a2ae9ed | 42 | |
6d718b9b TL |
43 | * For security reasons, access to host resources needs to be restricted. |
44 | Containers run in their own separate namespaces. Additionally some syscalls | |
45 | are not allowed within containers. | |
4a2ae9ed | 46 | |
6d718b9b TL |
47 | {pve} uses https://linuxcontainers.org/[Linux Containers (LXC)] as underlying |
48 | container technology. The ``Proxmox Container Toolkit'' (`pct`) simplifies the | |
49 | usage and management of LXC containers. | |
4a2ae9ed | 50 | |
14e97811 OB |
51 | Containers are tightly integrated with {pve}. This means that they are aware of |
52 | the cluster setup, and they can use the same network and storage resources as | |
53 | virtual machines. You can also use the {pve} firewall, or manage containers | |
54 | using the HA framework. | |
4a2ae9ed | 55 | |
6d718b9b TL |
56 | Our primary goal is to offer an environment as one would get from a VM, but |
57 | without the additional overhead. We call this ``System Containers''. | |
4a2ae9ed | 58 | |
6d718b9b | 59 | NOTE: If you want to run micro-containers, for example, 'Docker' or 'rkt', it |
70a42028 | 60 | is best to run them inside a VM. |
4a2ae9ed DM |
61 | |
62 | ||
99f6ae1a DM |
63 | Technology Overview |
64 | ------------------- | |
65 | ||
66 | * LXC (https://linuxcontainers.org/) | |
67 | ||
6d718b9b | 68 | * Integrated into {pve} graphical web user interface (GUI) |
99f6ae1a DM |
69 | |
70 | * Easy to use command line tool `pct` | |
71 | ||
72 | * Access via {pve} REST API | |
73 | ||
6d718b9b | 74 | * 'lxcfs' to provide containerized /proc file system |
99f6ae1a | 75 | |
6d718b9b | 76 | * Control groups ('cgroups') for resource isolation and limitation |
99f6ae1a | 77 | |
6d718b9b | 78 | * 'AppArmor' and 'seccomp' to improve security |
99f6ae1a | 79 | |
14e97811 | 80 | * Modern Linux kernels |
99f6ae1a DM |
81 | |
82 | * Image based deployment (templates) | |
83 | ||
6d718b9b | 84 | * Uses {pve} xref:chapter_storage[storage library] |
99f6ae1a | 85 | |
14e97811 | 86 | * Container setup from host (network, DNS, storage, etc.) |
99f6ae1a | 87 | |
69ab602f | 88 | |
80c0adcb | 89 | [[pct_container_images]] |
d61bab51 DM |
90 | Container Images |
91 | ---------------- | |
92 | ||
8c1189b6 | 93 | Container images, sometimes also referred to as ``templates'' or |
69ab602f | 94 | ``appliances'', are `tar` archives which contain everything to run a container. |
d61bab51 | 95 | |
69ab602f TL |
96 | {pve} itself provides a variety of basic templates for the most common Linux |
97 | distributions. They can be downloaded using the GUI or the `pveam` (short for | |
98 | {pve} Appliance Manager) command line utility. | |
99 | Additionally, https://www.turnkeylinux.org/[TurnKey Linux] container templates | |
100 | are also available to download. | |
d61bab51 | 101 | |
2a368b1e TL |
102 | The list of available templates is updated daily through the 'pve-daily-update' |
103 | timer. You can also trigger an update manually by executing: | |
3a6fa247 | 104 | |
14e97811 OB |
105 | ---- |
106 | # pveam update | |
107 | ---- | |
3a6fa247 | 108 | |
14e97811 | 109 | To view the list of available images run: |
3a6fa247 | 110 | |
14e97811 OB |
111 | ---- |
112 | # pveam available | |
113 | ---- | |
3a6fa247 | 114 | |
8c1189b6 FG |
115 | You can restrict this large list by specifying the `section` you are |
116 | interested in, for example basic `system` images: | |
3a6fa247 DM |
117 | |
118 | .List available system images | |
119 | ---- | |
120 | # pveam available --section system | |
14e97811 OB |
121 | system alpine-3.10-default_20190626_amd64.tar.xz |
122 | system alpine-3.9-default_20190224_amd64.tar.xz | |
123 | system archlinux-base_20190924-1_amd64.tar.gz | |
124 | system centos-6-default_20191016_amd64.tar.xz | |
125 | system centos-7-default_20190926_amd64.tar.xz | |
126 | system centos-8-default_20191016_amd64.tar.xz | |
127 | system debian-10.0-standard_10.0-1_amd64.tar.gz | |
128 | system debian-8.0-standard_8.11-1_amd64.tar.gz | |
129 | system debian-9.0-standard_9.7-1_amd64.tar.gz | |
130 | system fedora-30-default_20190718_amd64.tar.xz | |
131 | system fedora-31-default_20191029_amd64.tar.xz | |
132 | system gentoo-current-default_20190718_amd64.tar.xz | |
133 | system opensuse-15.0-default_20180907_amd64.tar.xz | |
134 | system opensuse-15.1-default_20190719_amd64.tar.xz | |
135 | system ubuntu-16.04-standard_16.04.5-1_amd64.tar.gz | |
136 | system ubuntu-18.04-standard_18.04.1-1_amd64.tar.gz | |
137 | system ubuntu-19.04-standard_19.04-1_amd64.tar.gz | |
138 | system ubuntu-19.10-standard_19.10-1_amd64.tar.gz | |
3a6fa247 DM |
139 | ---- |
140 | ||
69ab602f | 141 | Before you can use such a template, you need to download them into one of your |
2a368b1e TL |
142 | storages. If you're unsure to which one, you can simply use the `local` named |
143 | storage for that purpose. For clustered installations, it is preferred to use a | |
144 | shared storage so that all nodes can access those images. | |
3a6fa247 | 145 | |
14e97811 OB |
146 | ---- |
147 | # pveam download local debian-10.0-standard_10.0-1_amd64.tar.gz | |
148 | ---- | |
3a6fa247 | 149 | |
69ab602f TL |
150 | You are now ready to create containers using that image, and you can list all |
151 | downloaded images on storage `local` with: | |
24f73a63 DM |
152 | |
153 | ---- | |
154 | # pveam list local | |
14e97811 | 155 | local:vztmpl/debian-10.0-standard_10.0-1_amd64.tar.gz 219.95MB |
24f73a63 DM |
156 | ---- |
157 | ||
2a368b1e TL |
158 | TIP: You can also use the {pve} web interface GUI to download, list and delete |
159 | container templates. | |
160 | ||
161 | `pct` uses them to create a new container, for example: | |
162 | ||
163 | ---- | |
164 | # pct create 999 local:vztmpl/debian-10.0-standard_10.0-1_amd64.tar.gz | |
165 | ---- | |
166 | ||
69ab602f TL |
167 | The above command shows you the full {pve} volume identifiers. They include the |
168 | storage name, and most other {pve} commands can use them. For example you can | |
169 | delete that image later with: | |
24f73a63 | 170 | |
14e97811 OB |
171 | ---- |
172 | # pveam remove local:vztmpl/debian-10.0-standard_10.0-1_amd64.tar.gz | |
173 | ---- | |
d61bab51 | 174 | |
690cd737 | 175 | |
f3afbb70 | 176 | [[pct_settings]] |
4f785ca7 DM |
177 | Container Settings |
178 | ------------------ | |
179 | ||
304eb5a9 EK |
180 | [[pct_general]] |
181 | General Settings | |
182 | ~~~~~~~~~~~~~~~~ | |
183 | ||
1ff5e4e8 | 184 | [thumbnail="screenshot/gui-create-ct-general.png"] |
2225402c | 185 | |
304eb5a9 EK |
186 | General settings of a container include |
187 | ||
188 | * the *Node* : the physical server on which the container will run | |
69ab602f TL |
189 | * the *CT ID*: a unique number in this {pve} installation used to identify your |
190 | container | |
304eb5a9 EK |
191 | * *Hostname*: the hostname of the container |
192 | * *Resource Pool*: a logical group of containers and VMs | |
193 | * *Password*: the root password of the container | |
194 | * *SSH Public Key*: a public key for connecting to the root account over SSH | |
195 | * *Unprivileged container*: this option allows to choose at creation time | |
69ab602f | 196 | if you want to create a privileged or unprivileged container. |
304eb5a9 | 197 | |
14e97811 OB |
198 | Unprivileged Containers |
199 | ^^^^^^^^^^^^^^^^^^^^^^^ | |
200 | ||
69ab602f TL |
201 | Unprivileged containers use a new kernel feature called user namespaces. |
202 | The root UID 0 inside the container is mapped to an unprivileged user outside | |
203 | the container. This means that most security issues (container escape, resource | |
14e97811 OB |
204 | abuse, etc.) in these containers will affect a random unprivileged user, and |
205 | would be a generic kernel security bug rather than an LXC issue. The LXC team | |
206 | thinks unprivileged containers are safe by design. | |
207 | ||
208 | This is the default option when creating a new container. | |
209 | ||
69ab602f TL |
210 | NOTE: If the container uses systemd as an init system, please be aware the |
211 | systemd version running inside the container should be equal to or greater than | |
212 | 220. | |
14e97811 | 213 | |
304eb5a9 EK |
214 | |
215 | Privileged Containers | |
216 | ^^^^^^^^^^^^^^^^^^^^^ | |
217 | ||
c02ac25b TL |
218 | Security in containers is achieved by using mandatory access control 'AppArmor' |
219 | restrictions, 'seccomp' filters and Linux kernel namespaces. The LXC team | |
220 | considers this kind of container as unsafe, and they will not consider new | |
221 | container escape exploits to be security issues worthy of a CVE and quick fix. | |
222 | That's why privileged containers should only be used in trusted environments. | |
304eb5a9 | 223 | |
304eb5a9 | 224 | |
9a5e9443 | 225 | [[pct_cpu]] |
9a5e9443 DM |
226 | CPU |
227 | ~~~ | |
228 | ||
1ff5e4e8 | 229 | [thumbnail="screenshot/gui-create-ct-cpu.png"] |
097aa949 | 230 | |
14e97811 OB |
231 | You can restrict the number of visible CPUs inside the container using the |
232 | `cores` option. This is implemented using the Linux 'cpuset' cgroup | |
69ab602f TL |
233 | (**c**ontrol *group*). |
234 | A special task inside `pvestatd` tries to distribute running containers among | |
235 | available CPUs periodically. | |
236 | To view the assigned CPUs run the following command: | |
9a5e9443 DM |
237 | |
238 | ---- | |
239 | # pct cpusets | |
240 | --------------------- | |
241 | 102: 6 7 | |
242 | 105: 2 3 4 5 | |
243 | 108: 0 1 | |
244 | --------------------- | |
245 | ---- | |
246 | ||
14e97811 OB |
247 | Containers use the host kernel directly. All tasks inside a container are |
248 | handled by the host CPU scheduler. {pve} uses the Linux 'CFS' (**C**ompletely | |
249 | **F**air **S**cheduler) scheduler by default, which has additional bandwidth | |
250 | control options. | |
9a5e9443 DM |
251 | |
252 | [horizontal] | |
0725e3c6 | 253 | |
69ab602f TL |
254 | `cpulimit`: :: You can use this option to further limit assigned CPU time. |
255 | Please note that this is a floating point number, so it is perfectly valid to | |
256 | assign two cores to a container, but restrict overall CPU consumption to half a | |
257 | core. | |
9a5e9443 DM |
258 | + |
259 | ---- | |
260 | cores: 2 | |
261 | cpulimit: 0.5 | |
262 | ---- | |
263 | ||
69ab602f TL |
264 | `cpuunits`: :: This is a relative weight passed to the kernel scheduler. The |
265 | larger the number is, the more CPU time this container gets. Number is relative | |
266 | to the weights of all the other running containers. The default is 1024. You | |
267 | can use this setting to prioritize some containers. | |
9a5e9443 DM |
268 | |
269 | ||
270 | [[pct_memory]] | |
271 | Memory | |
272 | ~~~~~~ | |
273 | ||
1ff5e4e8 | 274 | [thumbnail="screenshot/gui-create-ct-memory.png"] |
097aa949 | 275 | |
9a5e9443 DM |
276 | Container memory is controlled using the cgroup memory controller. |
277 | ||
278 | [horizontal] | |
279 | ||
69ab602f TL |
280 | `memory`: :: Limit overall memory usage. This corresponds to the |
281 | `memory.limit_in_bytes` cgroup setting. | |
9a5e9443 | 282 | |
69ab602f TL |
283 | `swap`: :: Allows the container to use additional swap memory from the host |
284 | swap space. This corresponds to the `memory.memsw.limit_in_bytes` cgroup | |
285 | setting, which is set to the sum of both value (`memory + swap`). | |
9a5e9443 | 286 | |
4f785ca7 DM |
287 | |
288 | [[pct_mount_points]] | |
9e44e493 DM |
289 | Mount Points |
290 | ~~~~~~~~~~~~ | |
eeecce95 | 291 | |
1ff5e4e8 | 292 | [thumbnail="screenshot/gui-create-ct-root-disk.png"] |
097aa949 | 293 | |
14e97811 | 294 | The root mount point is configured with the `rootfs` property. You can |
69ab602f TL |
295 | configure up to 256 additional mount points. The corresponding options are |
296 | called `mp0` to `mp255`. They can contain the following settings: | |
01639994 FG |
297 | |
298 | include::pct-mountpoint-opts.adoc[] | |
299 | ||
69ab602f TL |
300 | Currently there are three types of mount points: storage backed mount points, |
301 | bind mounts, and device mounts. | |
9e44e493 | 302 | |
5eba0743 | 303 | .Typical container `rootfs` configuration |
4c3b5c77 DM |
304 | ---- |
305 | rootfs: thin1:base-100-disk-1,size=8G | |
306 | ---- | |
307 | ||
308 | ||
5eba0743 | 309 | Storage Backed Mount Points |
4c3b5c77 | 310 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
01639994 | 311 | |
9e44e493 | 312 | Storage backed mount points are managed by the {pve} storage subsystem and come |
eeecce95 WB |
313 | in three different flavors: |
314 | ||
5eba0743 | 315 | - Image based: these are raw images containing a single ext4 formatted file |
eeecce95 | 316 | system. |
5eba0743 | 317 | - ZFS subvolumes: these are technically bind mounts, but with managed storage, |
eeecce95 WB |
318 | and thus allow resizing and snapshotting. |
319 | - Directories: passing `size=0` triggers a special case where instead of a raw | |
320 | image a directory is created. | |
321 | ||
03782251 FG |
322 | NOTE: The special option syntax `STORAGE_ID:SIZE_IN_GB` for storage backed |
323 | mount point volumes will automatically allocate a volume of the specified size | |
69ab602f TL |
324 | on the specified storage. For example, calling |
325 | ||
326 | ---- | |
327 | pct set 100 -mp0 thin1:10,mp=/path/in/container | |
328 | ---- | |
329 | ||
330 | will allocate a 10GB volume on the storage `thin1` and replace the volume ID | |
331 | place holder `10` with the allocated volume ID, and setup the moutpoint in the | |
332 | container at `/path/in/container` | |
03782251 | 333 | |
4c3b5c77 | 334 | |
5eba0743 | 335 | Bind Mount Points |
4c3b5c77 | 336 | ^^^^^^^^^^^^^^^^^ |
01639994 | 337 | |
9baca183 FG |
338 | Bind mounts allow you to access arbitrary directories from your Proxmox VE host |
339 | inside a container. Some potential use cases are: | |
340 | ||
341 | - Accessing your home directory in the guest | |
342 | - Accessing an USB device directory in the guest | |
acccc49b | 343 | - Accessing an NFS mount from the host in the guest |
9baca183 | 344 | |
eeecce95 | 345 | Bind mounts are considered to not be managed by the storage subsystem, so you |
9baca183 | 346 | cannot make snapshots or deal with quotas from inside the container. With |
eeecce95 | 347 | unprivileged containers you might run into permission problems caused by the |
9baca183 FG |
348 | user mapping and cannot use ACLs. |
349 | ||
8c1189b6 | 350 | NOTE: The contents of bind mount points are not backed up when using `vzdump`. |
eeecce95 | 351 | |
69ab602f TL |
352 | WARNING: For security reasons, bind mounts should only be established using |
353 | source directories especially reserved for this purpose, e.g., a directory | |
354 | hierarchy under `/mnt/bindmounts`. Never bind mount system directories like | |
355 | `/`, `/var` or `/etc` into a container - this poses a great security risk. | |
9baca183 FG |
356 | |
357 | NOTE: The bind mount source path must not contain any symlinks. | |
358 | ||
359 | For example, to make the directory `/mnt/bindmounts/shared` accessible in the | |
360 | container with ID `100` under the path `/shared`, use a configuration line like | |
8c1189b6 FG |
361 | `mp0: /mnt/bindmounts/shared,mp=/shared` in `/etc/pve/lxc/100.conf`. |
362 | Alternatively, use `pct set 100 -mp0 /mnt/bindmounts/shared,mp=/shared` to | |
9baca183 | 363 | achieve the same result. |
6b707f2c | 364 | |
4c3b5c77 | 365 | |
5eba0743 | 366 | Device Mount Points |
4c3b5c77 | 367 | ^^^^^^^^^^^^^^^^^^^ |
fe154a4f | 368 | |
7432d78e FG |
369 | Device mount points allow to mount block devices of the host directly into the |
370 | container. Similar to bind mounts, device mounts are not managed by {PVE}'s | |
371 | storage subsystem, but the `quota` and `acl` options will be honored. | |
372 | ||
373 | NOTE: Device mount points should only be used under special circumstances. In | |
374 | most cases a storage backed mount point offers the same performance and a lot | |
375 | more features. | |
376 | ||
69ab602f TL |
377 | NOTE: The contents of device mount points are not backed up when using |
378 | `vzdump`. | |
01639994 | 379 | |
4c3b5c77 | 380 | |
80c0adcb | 381 | [[pct_container_network]] |
f5c351f0 DM |
382 | Network |
383 | ~~~~~~~ | |
04c569f6 | 384 | |
1ff5e4e8 | 385 | [thumbnail="screenshot/gui-create-ct-network.png"] |
097aa949 | 386 | |
69ab602f TL |
387 | You can configure up to 10 network interfaces for a single container. |
388 | The corresponding options are called `net0` to `net9`, and they can contain the | |
389 | following setting: | |
bac8c385 DM |
390 | |
391 | include::pct-network-opts.adoc[] | |
04c569f6 DM |
392 | |
393 | ||
139a9019 DM |
394 | [[pct_startup_and_shutdown]] |
395 | Automatic Start and Shutdown of Containers | |
396 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
397 | ||
14e97811 OB |
398 | To automatically start a container when the host system boots, select the |
399 | option 'Start at boot' in the 'Options' panel of the container in the web | |
400 | interface or run the following command: | |
139a9019 | 401 | |
14e97811 OB |
402 | ---- |
403 | # pct set CTID -onboot 1 | |
404 | ---- | |
139a9019 | 405 | |
4dbeb548 DM |
406 | .Start and Shutdown Order |
407 | // use the screenshot from qemu - its the same | |
1ff5e4e8 | 408 | [thumbnail="screenshot/gui-qemu-edit-start-order.png"] |
4dbeb548 | 409 | |
69ab602f TL |
410 | If you want to fine tune the boot order of your containers, you can use the |
411 | following parameters: | |
139a9019 | 412 | |
69ab602f TL |
413 | * *Start/Shutdown order*: Defines the start order priority. For example, set it |
414 | to 1 if you want the CT to be the first to be started. (We use the reverse | |
415 | startup order for shutdown, so a container with a start order of 1 would be | |
416 | the last to be shut down) | |
417 | * *Startup delay*: Defines the interval between this container start and | |
418 | subsequent containers starts. For example, set it to 240 if you want to wait | |
419 | 240 seconds before starting other containers. | |
139a9019 | 420 | * *Shutdown timeout*: Defines the duration in seconds {pve} should wait |
69ab602f TL |
421 | for the container to be offline after issuing a shutdown command. |
422 | By default this value is set to 60, which means that {pve} will issue a | |
423 | shutdown request, wait 60s for the machine to be offline, and if after 60s | |
424 | the machine is still online will notify that the shutdown action failed. | |
139a9019 | 425 | |
69ab602f TL |
426 | Please note that containers without a Start/Shutdown order parameter will |
427 | always start after those where the parameter is set, and this parameter only | |
139a9019 DM |
428 | makes sense between the machines running locally on a host, and not |
429 | cluster-wide. | |
430 | ||
c2c8eb89 DC |
431 | Hookscripts |
432 | ~~~~~~~~~~~ | |
433 | ||
434 | You can add a hook script to CTs with the config property `hookscript`. | |
435 | ||
14e97811 OB |
436 | ---- |
437 | # pct set 100 -hookscript local:snippets/hookscript.pl | |
438 | ---- | |
c2c8eb89 | 439 | |
69ab602f TL |
440 | It will be called during various phases of the guests lifetime. For an example |
441 | and documentation see the example script under | |
c2c8eb89 | 442 | `/usr/share/pve-docs/examples/guest-example-hookscript.pl`. |
139a9019 | 443 | |
bf7f598a TL |
444 | Security Considerations |
445 | ----------------------- | |
446 | ||
447 | Containers use the kernel of the host system. This exposes an attack surface | |
448 | for malicious users. In general, full virtual machines provide better | |
656d8b21 | 449 | isolation. This should be considered if containers are provided to unknown or |
bf7f598a TL |
450 | untrusted people. |
451 | ||
452 | To reduce the attack surface, LXC uses many security features like AppArmor, | |
453 | CGroups and kernel namespaces. | |
454 | ||
c02ac25b TL |
455 | AppArmor |
456 | ~~~~~~~~ | |
457 | ||
bf7f598a TL |
458 | AppArmor profiles are used to restrict access to possibly dangerous actions. |
459 | Some system calls, i.e. `mount`, are prohibited from execution. | |
460 | ||
461 | To trace AppArmor activity, use: | |
462 | ||
463 | ---- | |
464 | # dmesg | grep apparmor | |
465 | ---- | |
466 | ||
c02ac25b TL |
467 | Although it is not recommended, AppArmor can be disabled for a container. This |
468 | brings security risks with it. Some syscalls can lead to privilege escalation | |
469 | when executed within a container if the system is misconfigured or if a LXC or | |
470 | Linux Kernel vulnerability exists. | |
471 | ||
472 | To disable AppArmor for a container, add the following line to the container | |
473 | configuration file located at `/etc/pve/lxc/CTID.conf`: | |
474 | ||
475 | ---- | |
76aaaeab | 476 | lxc.apparmor.profile = unconfined |
c02ac25b TL |
477 | ---- |
478 | ||
479 | WARNING: Please note that this is not recommended for production use. | |
480 | ||
481 | ||
482 | // TODO: describe cgroups + seccomp a bit more. | |
483 | // TODO: pve-lxc-syscalld | |
484 | ||
485 | ||
0892a2c2 TL |
486 | Guest Operating System Configuration |
487 | ------------------------------------ | |
488 | ||
489 | {pve} tries to detect the Linux distribution in the container, and modifies | |
490 | some files. Here is a short list of things done at container startup: | |
491 | ||
492 | set /etc/hostname:: to set the container name | |
493 | ||
494 | modify /etc/hosts:: to allow lookup of the local hostname | |
495 | ||
496 | network setup:: pass the complete network setup to the container | |
497 | ||
498 | configure DNS:: pass information about DNS servers | |
499 | ||
500 | adapt the init system:: for example, fix the number of spawned getty processes | |
501 | ||
502 | set the root password:: when creating a new container | |
503 | ||
504 | rewrite ssh_host_keys:: so that each container has unique keys | |
505 | ||
506 | randomize crontab:: so that cron does not start at the same time on all containers | |
507 | ||
508 | Changes made by {PVE} are enclosed by comment markers: | |
509 | ||
510 | ---- | |
511 | # --- BEGIN PVE --- | |
512 | <data> | |
513 | # --- END PVE --- | |
514 | ---- | |
515 | ||
516 | Those markers will be inserted at a reasonable location in the file. If such a | |
517 | section already exists, it will be updated in place and will not be moved. | |
518 | ||
519 | Modification of a file can be prevented by adding a `.pve-ignore.` file for it. | |
520 | For instance, if the file `/etc/.pve-ignore.hosts` exists then the `/etc/hosts` | |
521 | file will not be touched. This can be a simple empty file created via: | |
522 | ||
523 | ---- | |
524 | # touch /etc/.pve-ignore.hosts | |
525 | ---- | |
526 | ||
527 | Most modifications are OS dependent, so they differ between different | |
528 | distributions and versions. You can completely disable modifications by | |
529 | manually setting the `ostype` to `unmanaged`. | |
530 | ||
531 | OS type detection is done by testing for certain files inside the | |
3d5c55fc TL |
532 | container. {pve} first checks the `/etc/os-release` file |
533 | footnote:[/etc/os-release replaces the multitude of per-distribution | |
534 | release files https://manpages.debian.org/stable/systemd/os-release.5.en.html]. | |
535 | If that file is not present, or it does not contain a clearly recognizable | |
536 | distribution identifier the following distribution specific release files are | |
537 | checked. | |
0892a2c2 TL |
538 | |
539 | Ubuntu:: inspect /etc/lsb-release (`DISTRIB_ID=Ubuntu`) | |
540 | ||
541 | Debian:: test /etc/debian_version | |
542 | ||
543 | Fedora:: test /etc/fedora-release | |
544 | ||
545 | RedHat or CentOS:: test /etc/redhat-release | |
546 | ||
547 | ArchLinux:: test /etc/arch-release | |
548 | ||
549 | Alpine:: test /etc/alpine-release | |
550 | ||
551 | Gentoo:: test /etc/gentoo-release | |
552 | ||
553 | NOTE: Container start fails if the configured `ostype` differs from the auto | |
554 | detected type. | |
555 | ||
556 | ||
b0df9949 TL |
557 | [[pct_container_storage]] |
558 | Container Storage | |
559 | ----------------- | |
560 | ||
561 | The {pve} LXC container storage model is more flexible than traditional | |
562 | container storage models. A container can have multiple mount points. This | |
563 | makes it possible to use the best suited storage for each application. | |
564 | ||
565 | For example the root file system of the container can be on slow and cheap | |
566 | storage while the database can be on fast and distributed storage via a second | |
567 | mount point. See section <<pct_mount_points, Mount Points>> for further | |
568 | details. | |
569 | ||
570 | Any storage type supported by the {pve} storage library can be used. This means | |
571 | that containers can be stored on local (for example `lvm`, `zfs` or directory), | |
572 | shared external (like `iSCSI`, `NFS`) or even distributed storage systems like | |
573 | Ceph. Advanced storage features like snapshots or clones can be used if the | |
574 | underlying storage supports them. The `vzdump` backup tool can use snapshots to | |
575 | provide consistent container backups. | |
576 | ||
577 | Furthermore, local devices or local directories can be mounted directly using | |
578 | 'bind mounts'. This gives access to local resources inside a container with | |
579 | practically zero overhead. Bind mounts can be used as an easy way to share data | |
580 | between containers. | |
581 | ||
582 | ||
583 | FUSE Mounts | |
584 | ~~~~~~~~~~~ | |
585 | ||
586 | WARNING: Because of existing issues in the Linux kernel's freezer subsystem the | |
587 | usage of FUSE mounts inside a container is strongly advised against, as | |
588 | containers need to be frozen for suspend or snapshot mode backups. | |
589 | ||
590 | If FUSE mounts cannot be replaced by other mounting mechanisms or storage | |
591 | technologies, it is possible to establish the FUSE mount on the Proxmox host | |
592 | and use a bind mount point to make it accessible inside the container. | |
593 | ||
594 | ||
595 | Using Quotas Inside Containers | |
596 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
597 | ||
598 | Quotas allow to set limits inside a container for the amount of disk space that | |
599 | each user can use. | |
600 | ||
601 | NOTE: This only works on ext4 image based storage types and currently only | |
602 | works with privileged containers. | |
603 | ||
604 | Activating the `quota` option causes the following mount options to be used for | |
605 | a mount point: | |
606 | `usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0` | |
607 | ||
608 | This allows quotas to be used like on any other system. You can initialize the | |
609 | `/aquota.user` and `/aquota.group` files by running: | |
610 | ||
611 | ---- | |
612 | # quotacheck -cmug / | |
613 | # quotaon / | |
614 | ---- | |
615 | ||
616 | Then edit the quotas using the `edquota` command. Refer to the documentation of | |
617 | the distribution running inside the container for details. | |
618 | ||
619 | NOTE: You need to run the above commands for every mount point by passing the | |
620 | mount point's path instead of just `/`. | |
621 | ||
622 | ||
623 | Using ACLs Inside Containers | |
624 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
625 | ||
626 | The standard Posix **A**ccess **C**ontrol **L**ists are also available inside | |
627 | containers. ACLs allow you to set more detailed file ownership than the | |
628 | traditional user/group/others model. | |
629 | ||
630 | ||
631 | Backup of Container mount points | |
632 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
633 | ||
634 | To include a mount point in backups, enable the `backup` option for it in the | |
635 | container configuration. For an existing mount point `mp0` | |
636 | ||
637 | ---- | |
638 | mp0: guests:subvol-100-disk-1,mp=/root/files,size=8G | |
639 | ---- | |
640 | ||
641 | add `backup=1` to enable it. | |
642 | ||
643 | ---- | |
644 | mp0: guests:subvol-100-disk-1,mp=/root/files,size=8G,backup=1 | |
645 | ---- | |
646 | ||
647 | NOTE: When creating a new mount point in the GUI, this option is enabled by | |
648 | default. | |
649 | ||
650 | To disable backups for a mount point, add `backup=0` in the way described | |
651 | above, or uncheck the *Backup* checkbox on the GUI. | |
652 | ||
653 | Replication of Containers mount points | |
654 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
655 | ||
656 | By default, additional mount points are replicated when the Root Disk is | |
657 | replicated. If you want the {pve} storage replication mechanism to skip a mount | |
658 | point, you can set the *Skip replication* option for that mount point. | |
659 | As of {pve} 5.0, replication requires a storage of type `zfspool`. Adding a | |
660 | mount point to a different type of storage when the container has replication | |
661 | configured requires to have *Skip replication* enabled for that mount point. | |
662 | ||
663 | ||
51e33128 FG |
664 | Backup and Restore |
665 | ------------------ | |
666 | ||
5eba0743 | 667 | |
2175e37b FG |
668 | Container Backup |
669 | ~~~~~~~~~~~~~~~~ | |
670 | ||
69ab602f TL |
671 | It is possible to use the `vzdump` tool for container backup. Please refer to |
672 | the `vzdump` manual page for details. | |
8c1189b6 | 673 | |
51e33128 | 674 | |
2175e37b FG |
675 | Restoring Container Backups |
676 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~ | |
677 | ||
69ab602f TL |
678 | Restoring container backups made with `vzdump` is possible using the `pct |
679 | restore` command. By default, `pct restore` will attempt to restore as much of | |
680 | the backed up container configuration as possible. It is possible to override | |
681 | the backed up configuration by manually setting container options on the | |
682 | command line (see the `pct` manual page for details). | |
2175e37b | 683 | |
8c1189b6 | 684 | NOTE: `pvesm extractconfig` can be used to view the backed up configuration |
2175e37b FG |
685 | contained in a vzdump archive. |
686 | ||
687 | There are two basic restore modes, only differing by their handling of mount | |
688 | points: | |
689 | ||
4c3b5c77 | 690 | |
8c1189b6 FG |
691 | ``Simple'' Restore Mode |
692 | ^^^^^^^^^^^^^^^^^^^^^^^ | |
2175e37b | 693 | |
69ab602f TL |
694 | If neither the `rootfs` parameter nor any of the optional `mpX` parameters are |
695 | explicitly set, the mount point configuration from the backed up configuration | |
696 | file is restored using the following steps: | |
2175e37b FG |
697 | |
698 | . Extract mount points and their options from backup | |
699 | . Create volumes for storage backed mount points (on storage provided with the | |
69ab602f | 700 | `storage` parameter, or default local storage if unset) |
2175e37b | 701 | . Extract files from backup archive |
69ab602f TL |
702 | . Add bind and device mount points to restored configuration (limited to root |
703 | user) | |
2175e37b FG |
704 | |
705 | NOTE: Since bind and device mount points are never backed up, no files are | |
706 | restored in the last step, but only the configuration options. The assumption | |
707 | is that such mount points are either backed up with another mechanism (e.g., | |
708 | NFS space that is bind mounted into many containers), or not intended to be | |
709 | backed up at all. | |
710 | ||
711 | This simple mode is also used by the container restore operations in the web | |
712 | interface. | |
713 | ||
4c3b5c77 | 714 | |
8c1189b6 FG |
715 | ``Advanced'' Restore Mode |
716 | ^^^^^^^^^^^^^^^^^^^^^^^^^ | |
2175e37b FG |
717 | |
718 | By setting the `rootfs` parameter (and optionally, any combination of `mpX` | |
8c1189b6 | 719 | parameters), the `pct restore` command is automatically switched into an |
2175e37b | 720 | advanced mode. This advanced mode completely ignores the `rootfs` and `mpX` |
69ab602f TL |
721 | configuration options contained in the backup archive, and instead only uses |
722 | the options explicitly provided as parameters. | |
2175e37b | 723 | |
69ab602f TL |
724 | This mode allows flexible configuration of mount point settings at restore |
725 | time, for example: | |
2175e37b FG |
726 | |
727 | * Set target storages, volume sizes and other options for each mount point | |
69ab602f | 728 | individually |
2175e37b FG |
729 | * Redistribute backed up files according to new mount point scheme |
730 | * Restore to device and/or bind mount points (limited to root user) | |
731 | ||
51e33128 | 732 | |
8c1189b6 | 733 | Managing Containers with `pct` |
04c569f6 DM |
734 | ------------------------------ |
735 | ||
6d718b9b TL |
736 | The ``Proxmox Container Toolkit'' (`pct`) is the command line tool to manage |
737 | {pve} containers. It enables you to create or destroy containers, as well as | |
738 | control the container execution (start, stop, reboot, migrate, etc.). It can be | |
739 | used to set parameters in the config file of a container, for example the | |
740 | network configuration or memory limits. | |
5eba0743 | 741 | |
04c569f6 DM |
742 | CLI Usage Examples |
743 | ~~~~~~~~~~~~~~~~~~ | |
744 | ||
69ab602f TL |
745 | Create a container based on a Debian template (provided you have already |
746 | downloaded the template via the web interface) | |
04c569f6 | 747 | |
14e97811 OB |
748 | ---- |
749 | # pct create 100 /var/lib/vz/template/cache/debian-10.0-standard_10.0-1_amd64.tar.gz | |
750 | ---- | |
04c569f6 DM |
751 | |
752 | Start container 100 | |
753 | ||
14e97811 OB |
754 | ---- |
755 | # pct start 100 | |
756 | ---- | |
04c569f6 DM |
757 | |
758 | Start a login session via getty | |
759 | ||
14e97811 OB |
760 | ---- |
761 | # pct console 100 | |
762 | ---- | |
04c569f6 DM |
763 | |
764 | Enter the LXC namespace and run a shell as root user | |
765 | ||
14e97811 OB |
766 | ---- |
767 | # pct enter 100 | |
768 | ---- | |
04c569f6 DM |
769 | |
770 | Display the configuration | |
771 | ||
14e97811 OB |
772 | ---- |
773 | # pct config 100 | |
774 | ---- | |
04c569f6 | 775 | |
69ab602f TL |
776 | Add a network interface called `eth0`, bridged to the host bridge `vmbr0`, set |
777 | the address and gateway, while it's running | |
04c569f6 | 778 | |
14e97811 OB |
779 | ---- |
780 | # pct set 100 -net0 name=eth0,bridge=vmbr0,ip=192.168.15.147/24,gw=192.168.15.1 | |
781 | ---- | |
04c569f6 DM |
782 | |
783 | Reduce the memory of the container to 512MB | |
784 | ||
14e97811 OB |
785 | ---- |
786 | # pct set 100 -memory 512 | |
787 | ---- | |
0585f29a | 788 | |
04c569f6 | 789 | |
fe57a420 FG |
790 | Obtaining Debugging Logs |
791 | ~~~~~~~~~~~~~~~~~~~~~~~~ | |
792 | ||
793 | In case `pct start` is unable to start a specific container, it might be | |
794 | helpful to collect debugging output by running `lxc-start` (replace `ID` with | |
795 | the container's ID): | |
796 | ||
14e97811 OB |
797 | ---- |
798 | # lxc-start -n ID -F -l DEBUG -o /tmp/lxc-ID.log | |
799 | ---- | |
fe57a420 | 800 | |
69ab602f TL |
801 | This command will attempt to start the container in foreground mode, to stop |
802 | the container run `pct shutdown ID` or `pct stop ID` in a second terminal. | |
fe57a420 FG |
803 | |
804 | The collected debug log is written to `/tmp/lxc-ID.log`. | |
805 | ||
806 | NOTE: If you have changed the container's configuration since the last start | |
807 | attempt with `pct start`, you need to run `pct start` at least once to also | |
808 | update the configuration used by `lxc-start`. | |
809 | ||
33f50e04 DC |
810 | [[pct_migration]] |
811 | Migration | |
812 | --------- | |
813 | ||
814 | If you have a cluster, you can migrate your Containers with | |
815 | ||
14e97811 OB |
816 | ---- |
817 | # pct migrate <ctid> <target> | |
818 | ---- | |
33f50e04 DC |
819 | |
820 | This works as long as your Container is offline. If it has local volumes or | |
14e97811 | 821 | mount points defined, the migration will copy the content over the network to |
ba021358 | 822 | the target host if the same storage is defined there. |
33f50e04 | 823 | |
656d8b21 | 824 | Running containers cannot live-migrated due to technical limitations. You can |
4c82550d TL |
825 | do a restart migration, which shuts down, moves and then starts a container |
826 | again on the target node. As containers are very lightweight, this results | |
827 | normally only in a downtime of some hundreds of milliseconds. | |
828 | ||
829 | A restart migration can be done through the web interface or by using the | |
830 | `--restart` flag with the `pct migrate` command. | |
33f50e04 | 831 | |
69ab602f TL |
832 | A restart migration will shut down the Container and kill it after the |
833 | specified timeout (the default is 180 seconds). Then it will migrate the | |
834 | Container like an offline migration and when finished, it starts the Container | |
835 | on the target node. | |
c7bc47af DM |
836 | |
837 | [[pct_configuration]] | |
838 | Configuration | |
839 | ------------- | |
840 | ||
69ab602f TL |
841 | The `/etc/pve/lxc/<CTID>.conf` file stores container configuration, where |
842 | `<CTID>` is the numeric ID of the given container. Like all other files stored | |
843 | inside `/etc/pve/`, they get automatically replicated to all other cluster | |
844 | nodes. | |
c7bc47af DM |
845 | |
846 | NOTE: CTIDs < 100 are reserved for internal purposes, and CTIDs need to be | |
847 | unique cluster wide. | |
848 | ||
849 | .Example Container Configuration | |
850 | ---- | |
851 | ostype: debian | |
852 | arch: amd64 | |
853 | hostname: www | |
854 | memory: 512 | |
855 | swap: 512 | |
856 | net0: bridge=vmbr0,hwaddr=66:64:66:64:64:36,ip=dhcp,name=eth0,type=veth | |
857 | rootfs: local:107/vm-107-disk-1.raw,size=7G | |
858 | ---- | |
859 | ||
69ab602f | 860 | The configuration files are simple text files. You can edit them using a normal |
da9679b6 | 861 | text editor, for example, `vi` or `nano`. |
69ab602f TL |
862 | This is sometimes useful to do small corrections, but keep in mind that you |
863 | need to restart the container to apply such changes. | |
c7bc47af | 864 | |
69ab602f TL |
865 | For that reason, it is usually better to use the `pct` command to generate and |
866 | modify those files, or do the whole thing using the GUI. | |
867 | Our toolkit is smart enough to instantaneously apply most changes to running | |
da9679b6 | 868 | containers. This feature is called ``hot plug'', and there is no need to restart |
69ab602f | 869 | the container in that case. |
c7bc47af | 870 | |
da9679b6 | 871 | In cases where a change cannot be hot-plugged, it will be registered as a |
69ab602f TL |
872 | pending change (shown in red color in the GUI). |
873 | They will only be applied after rebooting the container. | |
14e97811 | 874 | |
c7bc47af DM |
875 | |
876 | File Format | |
877 | ~~~~~~~~~~~ | |
878 | ||
69ab602f TL |
879 | The container configuration file uses a simple colon separated key/value |
880 | format. Each line has the following format: | |
c7bc47af DM |
881 | |
882 | ----- | |
883 | # this is a comment | |
884 | OPTION: value | |
885 | ----- | |
886 | ||
69ab602f TL |
887 | Blank lines in those files are ignored, and lines starting with a `#` character |
888 | are treated as comments and are also ignored. | |
c7bc47af | 889 | |
69ab602f | 890 | It is possible to add low-level, LXC style configuration directly, for example: |
c7bc47af | 891 | |
14e97811 OB |
892 | ---- |
893 | lxc.init_cmd: /sbin/my_own_init | |
894 | ---- | |
c7bc47af DM |
895 | |
896 | or | |
897 | ||
14e97811 OB |
898 | ---- |
899 | lxc.init_cmd = /sbin/my_own_init | |
900 | ---- | |
c7bc47af | 901 | |
14e97811 | 902 | The settings are passed directly to the LXC low-level tools. |
c7bc47af DM |
903 | |
904 | ||
905 | [[pct_snapshots]] | |
906 | Snapshots | |
907 | ~~~~~~~~~ | |
908 | ||
69ab602f TL |
909 | When you create a snapshot, `pct` stores the configuration at snapshot time |
910 | into a separate snapshot section within the same configuration file. For | |
911 | example, after creating a snapshot called ``testsnapshot'', your configuration | |
912 | file will look like this: | |
c7bc47af DM |
913 | |
914 | .Container configuration with snapshot | |
915 | ---- | |
916 | memory: 512 | |
917 | swap: 512 | |
918 | parent: testsnaphot | |
919 | ... | |
920 | ||
921 | [testsnaphot] | |
922 | memory: 512 | |
923 | swap: 512 | |
924 | snaptime: 1457170803 | |
925 | ... | |
926 | ---- | |
927 | ||
69ab602f TL |
928 | There are a few snapshot related properties like `parent` and `snaptime`. The |
929 | `parent` property is used to store the parent/child relationship between | |
930 | snapshots. `snaptime` is the snapshot creation time stamp (Unix epoch). | |
c7bc47af DM |
931 | |
932 | ||
933 | [[pct_options]] | |
934 | Options | |
935 | ~~~~~~~ | |
936 | ||
937 | include::pct.conf.5-opts.adoc[] | |
938 | ||
939 | ||
2a11aa70 DM |
940 | Locks |
941 | ----- | |
942 | ||
69ab602f TL |
943 | Container migrations, snapshots and backups (`vzdump`) set a lock to prevent |
944 | incompatible concurrent actions on the affected container. Sometimes you need | |
945 | to remove such a lock manually (e.g., after a power failure). | |
2a11aa70 | 946 | |
14e97811 OB |
947 | ---- |
948 | # pct unlock <CTID> | |
949 | ---- | |
2a11aa70 | 950 | |
69ab602f TL |
951 | CAUTION: Only do this if you are sure the action which set the lock is no |
952 | longer running. | |
2a11aa70 | 953 | |
fe57a420 | 954 | |
0c6b782f | 955 | ifdef::manvolnum[] |
3bd9d0cf DM |
956 | |
957 | Files | |
958 | ------ | |
959 | ||
960 | `/etc/pve/lxc/<CTID>.conf`:: | |
961 | ||
962 | Configuration file for the container '<CTID>'. | |
963 | ||
964 | ||
0c6b782f DM |
965 | include::pve-copyright.adoc[] |
966 | endif::manvolnum[] |