Precise certificate generation
[pve-docs.git] / qm.adoc
1 [[chapter_virtual_machines]]
2 ifdef::manvolnum[]
3 qm(1)
4 =====
5 :pve-toplevel:
8 ----
10 qm - Qemu/KVM Virtual Machine Manager
14 --------
16 include::qm.1-synopsis.adoc[]
19 -----------
20 endif::manvolnum[]
21 ifndef::manvolnum[]
22 Qemu/KVM Virtual Machines
23 =========================
24 :pve-toplevel:
25 endif::manvolnum[]
27 // deprecates
28 //
29 //
30 //
32 Qemu (short form for Quick Emulator) is an open source hypervisor that emulates a
33 physical computer. From the perspective of the host system where Qemu is
34 running, Qemu is a user program which has access to a number of local resources
35 like partitions, files, network cards which are then passed to an
36 emulated computer which sees them as if they were real devices.
38 A guest operating system running in the emulated computer accesses these
39 devices, and runs as it were running on real hardware. For instance you can pass
40 an iso image as a parameter to Qemu, and the OS running in the emulated computer
41 will see a real CDROM inserted in a CD drive.
43 Qemu can emulate a great variety of hardware from ARM to Sparc, but {pve} is
44 only concerned with 32 and 64 bits PC clone emulation, since it represents the
45 overwhelming majority of server hardware. The emulation of PC clones is also one
46 of the fastest due to the availability of processor extensions which greatly
47 speed up Qemu when the emulated architecture is the same as the host
48 architecture.
50 NOTE: You may sometimes encounter the term _KVM_ (Kernel-based Virtual Machine).
51 It means that Qemu is running with the support of the virtualization processor
52 extensions, via the Linux kvm module. In the context of {pve} _Qemu_ and
53 _KVM_ can be used interchangeably as Qemu in {pve} will always try to load the kvm
54 module.
56 Qemu inside {pve} runs as a root process, since this is required to access block
57 and PCI devices.
60 Emulated devices and paravirtualized devices
61 --------------------------------------------
63 The PC hardware emulated by Qemu includes a mainboard, network controllers,
64 scsi, ide and sata controllers, serial ports (the complete list can be seen in
65 the `kvm(1)` man page) all of them emulated in software. All these devices
66 are the exact software equivalent of existing hardware devices, and if the OS
67 running in the guest has the proper drivers it will use the devices as if it
68 were running on real hardware. This allows Qemu to runs _unmodified_ operating
69 systems.
71 This however has a performance cost, as running in software what was meant to
72 run in hardware involves a lot of extra work for the host CPU. To mitigate this,
73 Qemu can present to the guest operating system _paravirtualized devices_, where
74 the guest OS recognizes it is running inside Qemu and cooperates with the
75 hypervisor.
77 Qemu relies on the virtio virtualization standard, and is thus able to present
78 paravirtualized virtio devices, which includes a paravirtualized generic disk
79 controller, a paravirtualized network card, a paravirtualized serial port,
80 a paravirtualized SCSI controller, etc ...
82 It is highly recommended to use the virtio devices whenever you can, as they
83 provide a big performance improvement. Using the virtio generic disk controller
84 versus an emulated IDE controller will double the sequential write throughput,
85 as measured with `bonnie++(8)`. Using the virtio network interface can deliver
86 up to three times the throughput of an emulated Intel E1000 network card, as
87 measured with `iperf(1)`. footnote:[See this benchmark on the KVM wiki
91 [[qm_virtual_machines_settings]]
92 Virtual Machines Settings
93 -------------------------
95 Generally speaking {pve} tries to choose sane defaults for virtual machines
96 (VM). Make sure you understand the meaning of the settings you change, as it
97 could incur a performance slowdown, or putting your data at risk.
100 [[qm_general_settings]]
101 General Settings
102 ~~~~~~~~~~~~~~~~
104 [thumbnail="screenshot/gui-create-vm-general.png"]
106 General settings of a VM include
108 * the *Node* : the physical server on which the VM will run
109 * the *VM ID*: a unique number in this {pve} installation used to identify your VM
110 * *Name*: a free form text string you can use to describe the VM
111 * *Resource Pool*: a logical group of VMs
114 [[qm_os_settings]]
115 OS Settings
116 ~~~~~~~~~~~
118 [thumbnail="screenshot/gui-create-vm-os.png"]
120 When creating a VM, setting the proper Operating System(OS) allows {pve} to
121 optimize some low level parameters. For instance Windows OS expect the BIOS
122 clock to use the local time, while Unix based OS expect the BIOS clock to have
123 the UTC time.
126 [[qm_hard_disk]]
127 Hard Disk
128 ~~~~~~~~~
130 Qemu can emulate a number of storage controllers:
132 * the *IDE* controller, has a design which goes back to the 1984 PC/AT disk
133 controller. Even if this controller has been superseded by recent designs,
134 each and every OS you can think of has support for it, making it a great choice
135 if you want to run an OS released before 2003. You can connect up to 4 devices
136 on this controller.
138 * the *SATA* (Serial ATA) controller, dating from 2003, has a more modern
139 design, allowing higher throughput and a greater number of devices to be
140 connected. You can connect up to 6 devices on this controller.
142 * the *SCSI* controller, designed in 1985, is commonly found on server grade
143 hardware, and can connect up to 14 storage devices. {pve} emulates by default a
144 LSI 53C895A controller.
145 +
146 A SCSI controller of type _VirtIO SCSI_ is the recommended setting if you aim for
147 performance and is automatically selected for newly created Linux VMs since
148 {pve} 4.3. Linux distributions have support for this controller since 2012, and
149 FreeBSD since 2014. For Windows OSes, you need to provide an extra iso
150 containing the drivers during the installation.
151 //
152 If you aim at maximum performance, you can select a SCSI controller of type
153 _VirtIO SCSI single_ which will allow you to select the *IO Thread* option.
154 When selecting _VirtIO SCSI single_ Qemu will create a new controller for
155 each disk, instead of adding all disks to the same controller.
157 * The *VirtIO Block* controller, often just called VirtIO or virtio-blk,
158 is an older type of paravirtualized controller. It has been superseded by the
159 VirtIO SCSI Controller, in terms of features.
161 [thumbnail="screenshot/gui-create-vm-hard-disk.png"]
162 On each controller you attach a number of emulated hard disks, which are backed
163 by a file or a block device residing in the configured storage. The choice of
164 a storage type will determine the format of the hard disk image. Storages which
165 present block devices (LVM, ZFS, Ceph) will require the *raw disk image format*,
166 whereas files based storages (Ext4, NFS, CIFS, GlusterFS) will let you to choose
167 either the *raw disk image format* or the *QEMU image format*.
169 * the *QEMU image format* is a copy on write format which allows snapshots, and
170 thin provisioning of the disk image.
171 * the *raw disk image* is a bit-to-bit image of a hard disk, similar to what
172 you would get when executing the `dd` command on a block device in Linux. This
173 format does not support thin provisioning or snapshots by itself, requiring
174 cooperation from the storage layer for these tasks. It may, however, be up to
175 10% faster than the *QEMU image format*. footnote:[See this benchmark for details
177 * the *VMware image format* only makes sense if you intend to import/export the
178 disk image to other hypervisors.
180 Setting the *Cache* mode of the hard drive will impact how the host system will
181 notify the guest systems of block write completions. The *No cache* default
182 means that the guest system will be notified that a write is complete when each
183 block reaches the physical storage write queue, ignoring the host page cache.
184 This provides a good balance between safety and speed.
186 If you want the {pve} backup manager to skip a disk when doing a backup of a VM,
187 you can set the *No backup* option on that disk.
189 If you want the {pve} storage replication mechanism to skip a disk when starting
190 a replication job, you can set the *Skip replication* option on that disk.
191 As of {pve} 5.0, replication requires the disk images to be on a storage of type
192 `zfspool`, so adding a disk image to other storages when the VM has replication
193 configured requires to skip replication for this disk image.
195 If your storage supports _thin provisioning_ (see the storage chapter in the
196 {pve} guide), you can activate the *Discard* option on a drive. With *Discard*
197 set and a _TRIM_-enabled guest OS footnote:[TRIM, UNMAP, and discard
198], when the VM's filesystem
199 marks blocks as unused after deleting files, the controller will relay this
200 information to the storage, which will then shrink the disk image accordingly.
201 For the guest to be able to issue _TRIM_ commands, you must either use a
202 *VirtIO SCSI* (or *VirtIO SCSI Single*) controller or set the *SSD emulation*
203 option on the drive. Note that *Discard* is not supported on *VirtIO Block*
204 drives.
206 If you would like a drive to be presented to the guest as a solid-state drive
207 rather than a rotational hard disk, you can set the *SSD emulation* option on
208 that drive. There is no requirement that the underlying storage actually be
209 backed by SSDs; this feature can be used with physical media of any type.
210 Note that *SSD emulation* is not supported on *VirtIO Block* drives.
212 .IO Thread
213 The option *IO Thread* can only be used when using a disk with the
214 *VirtIO* controller, or with the *SCSI* controller, when the emulated controller
215 type is *VirtIO SCSI single*.
216 With this enabled, Qemu creates one I/O thread per storage controller,
217 instead of a single thread for all I/O, so it increases performance when
218 multiple disks are used and each disk has its own storage controller.
219 Note that backups do not currently work with *IO Thread* enabled.
222 [[qm_cpu]]
223 CPU
224 ~~~
226 [thumbnail="screenshot/gui-create-vm-cpu.png"]
228 A *CPU socket* is a physical slot on a PC motherboard where you can plug a CPU.
229 This CPU can then contain one or many *cores*, which are independent
230 processing units. Whether you have a single CPU socket with 4 cores, or two CPU
231 sockets with two cores is mostly irrelevant from a performance point of view.
232 However some software licenses depend on the number of sockets a machine has,
233 in that case it makes sense to set the number of sockets to what the license
234 allows you.
236 Increasing the number of virtual cpus (cores and sockets) will usually provide a
237 performance improvement though that is heavily dependent on the use of the VM.
238 Multithreaded applications will of course benefit from a large number of
239 virtual cpus, as for each virtual cpu you add, Qemu will create a new thread of
240 execution on the host system. If you're not sure about the workload of your VM,
241 it is usually a safe bet to set the number of *Total cores* to 2.
243 NOTE: It is perfectly safe if the _overall_ number of cores of all your VMs
244 is greater than the number of cores on the server (e.g., 4 VMs with each 4
245 cores on a machine with only 8 cores). In that case the host system will
246 balance the Qemu execution threads between your server cores, just like if you
247 were running a standard multithreaded application. However, {pve} will prevent
248 you from assigning more virtual CPU cores than physically available, as this will
249 only bring the performance down due to the cost of context switches.
251 [[qm_cpu_resource_limits]]
252 Resource Limits
253 ^^^^^^^^^^^^^^^
255 In addition to the number of virtual cores, you can configure how much resources
256 a VM can get in relation to the host CPU time and also in relation to other
257 VMs.
258 With the *cpulimit* (``Host CPU Time'') option you can limit how much CPU time
259 the whole VM can use on the host. It is a floating point value representing CPU
260 time in percent, so `1.0` is equal to `100%`, `2.5` to `250%` and so on. If a
261 single process would fully use one single core it would have `100%` CPU Time
262 usage. If a VM with four cores utilizes all its cores fully it would
263 theoretically use `400%`. In reality the usage may be even a bit higher as Qemu
264 can have additional threads for VM peripherals besides the vCPU core ones.
265 This setting can be useful if a VM should have multiple vCPUs, as it runs a few
266 processes in parallel, but the VM as a whole should not be able to run all
267 vCPUs at 100% at the same time. Using a specific example: lets say we have a VM
268 which would profit from having 8 vCPUs, but at no time all of those 8 cores
269 should run at full load - as this would make the server so overloaded that
270 other VMs and CTs would get to less CPU. So, we set the *cpulimit* limit to
271 `4.0` (=400%). If all cores do the same heavy work they would all get 50% of a
272 real host cores CPU time. But, if only 4 would do work they could still get
273 almost 100% of a real core each.
275 NOTE: VMs can, depending on their configuration, use additional threads e.g.,
276 for networking or IO operations but also live migration. Thus a VM can show up
277 to use more CPU time than just its virtual CPUs could use. To ensure that a VM
278 never uses more CPU time than virtual CPUs assigned set the *cpulimit* setting
279 to the same value as the total core count.
281 The second CPU resource limiting setting, *cpuunits* (nowadays often called CPU
282 shares or CPU weight), controls how much CPU time a VM gets in regards to other
283 VMs running. It is a relative weight which defaults to `1024`, if you increase
284 this for a VM it will be prioritized by the scheduler in comparison to other
285 VMs with lower weight. E.g., if VM 100 has set the default 1024 and VM 200 was
286 changed to `2048`, the latter VM 200 would receive twice the CPU bandwidth than
287 the first VM 100.
289 For more information see `man systemd.resource-control`, here `CPUQuota`
290 corresponds to `cpulimit` and `CPUShares` corresponds to our `cpuunits`
291 setting, visit its Notes section for references and implementation details.
293 CPU Type
294 ^^^^^^^^
296 Qemu can emulate a number different of *CPU types* from 486 to the latest Xeon
297 processors. Each new processor generation adds new features, like hardware
298 assisted 3d rendering, random number generation, memory protection, etc ...
299 Usually you should select for your VM a processor type which closely matches the
300 CPU of the host system, as it means that the host CPU features (also called _CPU
301 flags_ ) will be available in your VMs. If you want an exact match, you can set
302 the CPU type to *host* in which case the VM will have exactly the same CPU flags
303 as your host system.
305 This has a downside though. If you want to do a live migration of VMs between
306 different hosts, your VM might end up on a new system with a different CPU type.
307 If the CPU flags passed to the guest are missing, the qemu process will stop. To
308 remedy this Qemu has also its own CPU type *kvm64*, that {pve} uses by defaults.
309 kvm64 is a Pentium 4 look a like CPU type, which has a reduced CPU flags set,
310 but is guaranteed to work everywhere.
312 In short, if you care about live migration and moving VMs between nodes, leave
313 the kvm64 default. If you don’t care about live migration or have a homogeneous
314 cluster where all nodes have the same CPU, set the CPU type to host, as in
315 theory this will give your guests maximum performance.
317 Meltdown / Spectre related CPU flags
318 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
320 There are several CPU flags related to the Meltdown and Spectre vulnerabilities
321 footnote:[Meltdown Attack] which need to be set
322 manually unless the selected CPU type of your VM already enables them by default.
324 There are two requirements that need to be fulfilled in order to use these
325 CPU flags:
327 * The host CPU(s) must support the feature and propagate it to the guest's virtual CPU(s)
328 * The guest operating system must be updated to a version which mitigates the
329 attacks and is able to utilize the CPU feature
331 Otherwise you need to set the desired CPU flag of the virtual CPU, either by
332 editing the CPU options in the WebUI, or by setting the 'flags' property of the
333 'cpu' option in the VM configuration file.
335 For Spectre v1,v2,v4 fixes, your CPU or system vendor also needs to provide a
336 so-called ``microcode update'' footnote:[You can use `intel-microcode' /
337 `amd-microcode' from Debian non-free if your vendor does not provide such an
338 update. Note that not all affected CPUs can be updated to support spec-ctrl.]
339 for your CPU.
342 To check if the {pve} host is vulnerable, execute the following command as root:
344 ----
345 for f in /sys/devices/system/cpu/vulnerabilities/*; do echo "${f##*/} -" $(cat "$f"); done
346 ----
348 A community script is also available to detect is the host is still vulnerable.
349 footnote:[spectre-meltdown-checker]
351 Intel processors
352 ^^^^^^^^^^^^^^^^
354 * 'pcid'
355 +
356 This reduces the performance impact of the Meltdown (CVE-2017-5754) mitigation
357 called 'Kernel Page-Table Isolation (KPTI)', which effectively hides
358 the Kernel memory from the user space. Without PCID, KPTI is quite an expensive
359 mechanism footnote:[PCID is now a critical performance/security feature on x86
361 +
362 To check if the {pve} host supports PCID, execute the following command as root:
363 +
364 ----
365 # grep ' pcid ' /proc/cpuinfo
366 ----
367 +
368 If this does not return empty your host's CPU has support for 'pcid'.
370 * 'spec-ctrl'
371 +
372 Required to enable the Spectre v1 (CVE-2017-5753) and Spectre v2 (CVE-2017-5715) fix,
373 in cases where retpolines are not sufficient.
374 Included by default in Intel CPU models with -IBRS suffix.
375 Must be explicitly turned on for Intel CPU models without -IBRS suffix.
376 Requires an updated host CPU microcode (intel-microcode >= 20180425).
377 +
378 * 'ssbd'
379 +
380 Required to enable the Spectre V4 (CVE-2018-3639) fix. Not included by default in any Intel CPU model.
381 Must be explicitly turned on for all Intel CPU models.
382 Requires an updated host CPU microcode(intel-microcode >= 20180703).
385 AMD processors
386 ^^^^^^^^^^^^^^
388 * 'ibpb'
389 +
390 Required to enable the Spectre v1 (CVE-2017-5753) and Spectre v2 (CVE-2017-5715) fix,
391 in cases where retpolines are not sufficient.
392 Included by default in AMD CPU models with -IBPB suffix.
393 Must be explicitly turned on for AMD CPU models without -IBPB suffix.
394 Requires the host CPU microcode to support this feature before it can be used for guest CPUs.
398 * 'virt-ssbd'
399 +
400 Required to enable the Spectre v4 (CVE-2018-3639) fix.
401 Not included by default in any AMD CPU model.
402 Must be explicitly turned on for all AMD CPU models.
403 This should be provided to guests, even if amd-ssbd is also provided, for maximum guest compatibility.
404 Note that this must be explicitly enabled when when using the "host" cpu model,
405 because this is a virtual feature which does not exist in the physical CPUs.
408 * 'amd-ssbd'
409 +
410 Required to enable the Spectre v4 (CVE-2018-3639) fix.
411 Not included by default in any AMD CPU model. Must be explicitly turned on for all AMD CPU models.
412 This provides higher performance than virt-ssbd, therefore a host supporting this should always expose this to guests if possible.
413 virt-ssbd should none the less also be exposed for maximum guest compatibility as some kernels only know about virt-ssbd.
416 * 'amd-no-ssb'
417 +
418 Recommended to indicate the host is not vulnerable to Spectre V4 (CVE-2018-3639).
419 Not included by default in any AMD CPU model.
420 Future hardware generations of CPU will not be vulnerable to CVE-2018-3639,
421 and thus the guest should be told not to enable its mitigations, by exposing amd-no-ssb.
422 This is mutually exclusive with virt-ssbd and amd-ssbd.
425 NUMA
426 ^^^^
427 You can also optionally emulate a *NUMA*
428 footnote:[] architecture
429 in your VMs. The basics of the NUMA architecture mean that instead of having a
430 global memory pool available to all your cores, the memory is spread into local
431 banks close to each socket.
432 This can bring speed improvements as the memory bus is not a bottleneck
433 anymore. If your system has a NUMA architecture footnote:[if the command
434 `numactl --hardware | grep available` returns more than one node, then your host
435 system has a NUMA architecture] we recommend to activate the option, as this
436 will allow proper distribution of the VM resources on the host system.
437 This option is also required to hot-plug cores or RAM in a VM.
439 If the NUMA option is used, it is recommended to set the number of sockets to
440 the number of nodes of the host system.
442 vCPU hot-plug
443 ^^^^^^^^^^^^^
445 Modern operating systems introduced the capability to hot-plug and, to a
446 certain extent, hot-unplug CPUs in a running systems. Virtualisation allows us
447 to avoid a lot of the (physical) problems real hardware can cause in such
448 scenarios.
449 Still, this is a rather new and complicated feature, so its use should be
450 restricted to cases where its absolutely needed. Most of the functionality can
451 be replicated with other, well tested and less complicated, features, see
452 xref:qm_cpu_resource_limits[Resource Limits].
454 In {pve} the maximal number of plugged CPUs is always `cores * sockets`.
455 To start a VM with less than this total core count of CPUs you may use the
456 *vpus* setting, it denotes how many vCPUs should be plugged in at VM start.
458 Currently only this feature is only supported on Linux, a kernel newer than 3.10
459 is needed, a kernel newer than 4.7 is recommended.
461 You can use a udev rule as follow to automatically set new CPUs as online in
462 the guest:
464 ----
465 SUBSYSTEM=="cpu", ACTION=="add", TEST=="online", ATTR{online}=="0", ATTR{online}="1"
466 ----
468 Save this under /etc/udev/rules.d/ as a file ending in `.rules`.
470 Note: CPU hot-remove is machine dependent and requires guest cooperation.
471 The deletion command does not guarantee CPU removal to actually happen,
472 typically it's a request forwarded to guest using target dependent mechanism,
473 e.g., ACPI on x86/amd64.
476 [[qm_memory]]
477 Memory
478 ~~~~~~
480 For each VM you have the option to set a fixed size memory or asking
481 {pve} to dynamically allocate memory based on the current RAM usage of the
482 host.
484 .Fixed Memory Allocation
485 [thumbnail="screenshot/gui-create-vm-memory.png"]
487 When setting memory and minimum memory to the same amount
488 {pve} will simply allocate what you specify to your VM.
490 Even when using a fixed memory size, the ballooning device gets added to the
491 VM, because it delivers useful information such as how much memory the guest
492 really uses.
493 In general, you should leave *ballooning* enabled, but if you want to disable
494 it (e.g. for debugging purposes), simply uncheck
495 *Ballooning Device* or set
497 balloon: 0
499 in the configuration.
501 .Automatic Memory Allocation
503 // see autoballoon() in
504 When setting the minimum memory lower than memory, {pve} will make sure that the
505 minimum amount you specified is always available to the VM, and if RAM usage on
506 the host is below 80%, will dynamically add memory to the guest up to the
507 maximum memory specified.
509 When the host is running low on RAM, the VM will then release some memory
510 back to the host, swapping running processes if needed and starting the oom
511 killer in last resort. The passing around of memory between host and guest is
512 done via a special `balloon` kernel driver running inside the guest, which will
513 grab or release memory pages from the host.
514 footnote:[A good explanation of the inner workings of the balloon driver can be found here]
516 When multiple VMs use the autoallocate facility, it is possible to set a
517 *Shares* coefficient which indicates the relative amount of the free host memory
518 that each VM should take. Suppose for instance you have four VMs, three of them
519 running an HTTP server and the last one is a database server. To cache more
520 database blocks in the database server RAM, you would like to prioritize the
521 database VM when spare RAM is available. For this you assign a Shares property
522 of 3000 to the database VM, leaving the other VMs to the Shares default setting
523 of 1000. The host server has 32GB of RAM, and is currently using 16GB, leaving 32
524 * 80/100 - 16 = 9GB RAM to be allocated to the VMs. The database VM will get 9 *
525 3000 / (3000 + 1000 + 1000 + 1000) = 4.5 GB extra RAM and each HTTP server will
526 get 1.5 GB.
528 All Linux distributions released after 2010 have the balloon kernel driver
529 included. For Windows OSes, the balloon driver needs to be added manually and can
530 incur a slowdown of the guest, so we don't recommend using it on critical
531 systems.
532 // see
534 When allocating RAM to your VMs, a good rule of thumb is always to leave 1GB
535 of RAM available to the host.
538 [[qm_network_device]]
539 Network Device
540 ~~~~~~~~~~~~~~
542 [thumbnail="screenshot/gui-create-vm-network.png"]
544 Each VM can have many _Network interface controllers_ (NIC), of four different
545 types:
547 * *Intel E1000* is the default, and emulates an Intel Gigabit network card.
548 * the *VirtIO* paravirtualized NIC should be used if you aim for maximum
549 performance. Like all VirtIO devices, the guest OS should have the proper driver
550 installed.
551 * the *Realtek 8139* emulates an older 100 MB/s network card, and should
552 only be used when emulating older operating systems ( released before 2002 )
553 * the *vmxnet3* is another paravirtualized device, which should only be used
554 when importing a VM from another hypervisor.
556 {pve} will generate for each NIC a random *MAC address*, so that your VM is
557 addressable on Ethernet networks.
559 The NIC you added to the VM can follow one of two different models:
561 * in the default *Bridged mode* each virtual NIC is backed on the host by a
562 _tap device_, ( a software loopback device simulating an Ethernet NIC ). This
563 tap device is added to a bridge, by default vmbr0 in {pve}. In this mode, VMs
564 have direct access to the Ethernet LAN on which the host is located.
565 * in the alternative *NAT mode*, each virtual NIC will only communicate with
566 the Qemu user networking stack, where a built-in router and DHCP server can
567 provide network access. This built-in DHCP will serve addresses in the private
568 range. The NAT mode is much slower than the bridged mode, and
569 should only be used for testing. This mode is only available via CLI or the API,
570 but not via the WebUI.
572 You can also skip adding a network device when creating a VM by selecting *No
573 network device*.
575 .Multiqueue
576 If you are using the VirtIO driver, you can optionally activate the
577 *Multiqueue* option. This option allows the guest OS to process networking
578 packets using multiple virtual CPUs, providing an increase in the total number
579 of packets transferred.
581 //
582 When using the VirtIO driver with {pve}, each NIC network queue is passed to the
583 host kernel, where the queue will be processed by a kernel thread spawned by the
584 vhost driver. With this option activated, it is possible to pass _multiple_
585 network queues to the host kernel for each NIC.
587 //
588 When using Multiqueue, it is recommended to set it to a value equal
589 to the number of Total Cores of your guest. You also need to set in
590 the VM the number of multi-purpose channels on each VirtIO NIC with the ethtool
591 command:
593 `ethtool -L ens1 combined X`
595 where X is the number of the number of vcpus of the VM.
597 You should note that setting the Multiqueue parameter to a value greater
598 than one will increase the CPU load on the host and guest systems as the
599 traffic increases. We recommend to set this option only when the VM has to
600 process a great number of incoming connections, such as when the VM is running
601 as a router, reverse proxy or a busy HTTP server doing long polling.
603 [[qm_display]]
604 Display
605 ~~~~~~~
607 QEMU can virtualize a few types of VGA hardware. Some examples are:
609 * *std*, the default, emulates a card with Bochs VBE extensions.
610 * *cirrus*, this was once the default, it emulates a very old hardware module
611 with all its problems. This display type should only be used if really
612 necessary footnote:[
613 qemu: using cirrus considered harmful], e.g., if using Windows XP or earlier
614 * *vmware*, is a VMWare SVGA-II compatible adapter.
615 * *qxl*, is the QXL paravirtualized graphics card. Selecting this also
616 enables SPICE for the VM.
618 You can edit the amount of memory given to the virtual GPU, by setting
619 the 'memory' option. This can enable higher resolutions inside the VM,
620 especially with SPICE/QXL.
622 As the memory is reserved by display device, selecting Multi-Monitor mode
623 for SPICE (e.g., `qxl2` for dual monitors) has some implications:
625 * Windows needs a device for each monitor, so if your 'ostype' is some
626 version of Windows, {pve} gives the VM an extra device per monitor.
627 Each device gets the specified amount of memory.
629 * Linux VMs, can always enable more virtual monitors, but selecting
630 a Multi-Monitor mode multiplies the memory given to the device with
631 the number of monitors.
633 Selecting `serialX` as display 'type' disables the VGA output, and redirects
634 the Web Console to the selected serial port. A configured display 'memory'
635 setting will be ignored in that case.
637 [[qm_usb_passthrough]]
638 USB Passthrough
639 ~~~~~~~~~~~~~~~
641 There are two different types of USB passthrough devices:
643 * Host USB passthrough
644 * SPICE USB passthrough
646 Host USB passthrough works by giving a VM a USB device of the host.
647 This can either be done via the vendor- and product-id, or
648 via the host bus and port.
650 The vendor/product-id looks like this: *0123:abcd*,
651 where *0123* is the id of the vendor, and *abcd* is the id
652 of the product, meaning two pieces of the same usb device
653 have the same id.
655 The bus/port looks like this: *1-2.3.4*, where *1* is the bus
656 and *2.3.4* is the port path. This represents the physical
657 ports of your host (depending of the internal order of the
658 usb controllers).
660 If a device is present in a VM configuration when the VM starts up,
661 but the device is not present in the host, the VM can boot without problems.
662 As soon as the device/port is available in the host, it gets passed through.
664 WARNING: Using this kind of USB passthrough means that you cannot move
665 a VM online to another host, since the hardware is only available
666 on the host the VM is currently residing.
668 The second type of passthrough is SPICE USB passthrough. This is useful
669 if you use a SPICE client which supports it. If you add a SPICE USB port
670 to your VM, you can passthrough a USB device from where your SPICE client is,
671 directly to the VM (for example an input device or hardware dongle).
674 [[qm_bios_and_uefi]]
675 BIOS and UEFI
676 ~~~~~~~~~~~~~
678 In order to properly emulate a computer, QEMU needs to use a firmware.
679 By default QEMU uses *SeaBIOS* for this, which is an open-source, x86 BIOS
680 implementation. SeaBIOS is a good choice for most standard setups.
682 There are, however, some scenarios in which a BIOS is not a good firmware
683 to boot from, e.g. if you want to do VGA passthrough. footnote:[Alex Williamson has a very good blog entry about this.
685 In such cases, you should rather use *OVMF*, which is an open-source UEFI implementation. footnote:[See the OVMF Project]
687 If you want to use OVMF, there are several things to consider:
689 In order to save things like the *boot order*, there needs to be an EFI Disk.
690 This disk will be included in backups and snapshots, and there can only be one.
692 You can create such a disk with the following command:
694 qm set <vmid> -efidisk0 <storage>:1,format=<format>
696 Where *<storage>* is the storage where you want to have the disk, and
697 *<format>* is a format which the storage supports. Alternatively, you can
698 create such a disk through the web interface with 'Add' -> 'EFI Disk' in the
699 hardware section of a VM.
701 When using OVMF with a virtual display (without VGA passthrough),
702 you need to set the client resolution in the OVMF menu(which you can reach
703 with a press of the ESC button during boot), or you have to choose
704 SPICE as the display type.
706 [[qm_startup_and_shutdown]]
707 Automatic Start and Shutdown of Virtual Machines
708 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
710 After creating your VMs, you probably want them to start automatically
711 when the host system boots. For this you need to select the option 'Start at
712 boot' from the 'Options' Tab of your VM in the web interface, or set it with
713 the following command:
715 qm set <vmid> -onboot 1
717 .Start and Shutdown Order
719 [thumbnail="screenshot/gui-qemu-edit-start-order.png"]
721 In some case you want to be able to fine tune the boot order of your
722 VMs, for instance if one of your VM is providing firewalling or DHCP
723 to other guest systems. For this you can use the following
724 parameters:
726 * *Start/Shutdown order*: Defines the start order priority. E.g. set it to 1 if
727 you want the VM to be the first to be started. (We use the reverse startup
728 order for shutdown, so a machine with a start order of 1 would be the last to
729 be shut down). If multiple VMs have the same order defined on a host, they will
730 additionally be ordered by 'VMID' in ascending order.
731 * *Startup delay*: Defines the interval between this VM start and subsequent
732 VMs starts . E.g. set it to 240 if you want to wait 240 seconds before starting
733 other VMs.
734 * *Shutdown timeout*: Defines the duration in seconds {pve} should wait
735 for the VM to be offline after issuing a shutdown command.
736 By default this value is set to 180, which means that {pve} will issue a
737 shutdown request and wait 180 seconds for the machine to be offline. If
738 the machine is still online after the timeout it will be stopped forcefully.
740 NOTE: VMs managed by the HA stack do not follow the 'start on boot' and
741 'boot order' options currently. Those VMs will be skipped by the startup and
742 shutdown algorithm as the HA manager itself ensures that VMs get started and
743 stopped.
745 Please note that machines without a Start/Shutdown order parameter will always
746 start after those where the parameter is set. Further, this parameter can only
747 be enforced between virtual machines running on the same host, not
748 cluster-wide.
751 [[qm_migration]]
752 Migration
753 ---------
755 [thumbnail="screenshot/gui-qemu-migrate.png"]
757 If you have a cluster, you can migrate your VM to another host with
759 qm migrate <vmid> <target>
761 There are generally two mechanisms for this
763 * Online Migration (aka Live Migration)
764 * Offline Migration
766 Online Migration
767 ~~~~~~~~~~~~~~~~
769 When your VM is running and it has no local resources defined (such as disks
770 on local storage, passed through devices, etc.) you can initiate a live
771 migration with the -online flag.
773 How it works
774 ^^^^^^^^^^^^
776 This starts a Qemu Process on the target host with the 'incoming' flag, which
777 means that the process starts and waits for the memory data and device states
778 from the source Virtual Machine (since all other resources, e.g. disks,
779 are shared, the memory content and device state are the only things left
780 to transmit).
782 Once this connection is established, the source begins to send the memory
783 content asynchronously to the target. If the memory on the source changes,
784 those sections are marked dirty and there will be another pass of sending data.
785 This happens until the amount of data to send is so small that it can
786 pause the VM on the source, send the remaining data to the target and start
787 the VM on the target in under a second.
789 Requirements
790 ^^^^^^^^^^^^
792 For Live Migration to work, there are some things required:
794 * The VM has no local resources (e.g. passed through devices, local disks, etc.)
795 * The hosts are in the same {pve} cluster.
796 * The hosts have a working (and reliable) network connection.
797 * The target host must have the same or higher versions of the
798 {pve} packages. (It *might* work the other way, but this is never guaranteed)
800 Offline Migration
801 ~~~~~~~~~~~~~~~~~
803 If you have local resources, you can still offline migrate your VMs,
804 as long as all disk are on storages, which are defined on both hosts.
805 Then the migration will copy the disk over the network to the target host.
807 [[qm_copy_and_clone]]
808 Copies and Clones
809 -----------------
811 [thumbnail="screenshot/gui-qemu-full-clone.png"]
813 VM installation is usually done using an installation media (CD-ROM)
814 from the operation system vendor. Depending on the OS, this can be a
815 time consuming task one might want to avoid.
817 An easy way to deploy many VMs of the same type is to copy an existing
818 VM. We use the term 'clone' for such copies, and distinguish between
819 'linked' and 'full' clones.
821 Full Clone::
823 The result of such copy is an independent VM. The
824 new VM does not share any storage resources with the original.
825 +
827 It is possible to select a *Target Storage*, so one can use this to
828 migrate a VM to a totally different storage. You can also change the
829 disk image *Format* if the storage driver supports several formats.
830 +
832 NOTE: A full clone needs to read and copy all VM image data. This is
833 usually much slower than creating a linked clone.
834 +
836 Some storage types allows to copy a specific *Snapshot*, which
837 defaults to the 'current' VM data. This also means that the final copy
838 never includes any additional snapshots from the original VM.
841 Linked Clone::
843 Modern storage drivers support a way to generate fast linked
844 clones. Such a clone is a writable copy whose initial contents are the
845 same as the original data. Creating a linked clone is nearly
846 instantaneous, and initially consumes no additional space.
847 +
849 They are called 'linked' because the new image still refers to the
850 original. Unmodified data blocks are read from the original image, but
851 modification are written (and afterwards read) from a new
852 location. This technique is called 'Copy-on-write'.
853 +
855 This requires that the original volume is read-only. With {pve} one
856 can convert any VM into a read-only <<qm_templates, Template>>). Such
857 templates can later be used to create linked clones efficiently.
858 +
860 NOTE: You cannot delete an original template while linked clones
861 exist.
862 +
864 It is not possible to change the *Target storage* for linked clones,
865 because this is a storage internal feature.
868 The *Target node* option allows you to create the new VM on a
869 different node. The only restriction is that the VM is on shared
870 storage, and that storage is also available on the target node.
872 To avoid resource conflicts, all network interface MAC addresses get
873 randomized, and we generate a new 'UUID' for the VM BIOS (smbios1)
874 setting.
877 [[qm_templates]]
878 Virtual Machine Templates
879 -------------------------
881 One can convert a VM into a Template. Such templates are read-only,
882 and you can use them to create linked clones.
884 NOTE: It is not possible to start templates, because this would modify
885 the disk images. If you want to change the template, create a linked
886 clone and modify that.
888 VM Generation ID
889 ----------------
891 {pve} supports Virtual Machine Generation ID ('vmgenid') footnote:[Official
892 'vmgenid' Specification
894 for virtual machines.
895 This can be used by the guest operating system to detect any event resulting
896 in a time shift event, for example, restoring a backup or a snapshot rollback.
898 When creating new VMs, a 'vmgenid' will be automatically generated and saved
899 in its configuration file.
901 To create and add a 'vmgenid' to an already existing VM one can pass the
902 special value `1' to let {pve} autogenerate one or manually set the 'UUID'
903 footnote:[Online GUID generator] by using it as value,
904 e.g.:
906 ----
907 qm set VMID -vmgenid 1
908 qm set VMID -vmgenid 00000000-0000-0000-0000-000000000000
909 ----
911 NOTE: The initial addition of a 'vmgenid' device to an existing VM, may result
912 in the same effects as a change on snapshot rollback, backup restore, etc., has
913 as the VM can interpret this as generation change.
915 In the rare case the 'vmgenid' mechanism is not wanted one can pass `0' for
916 its value on VM creation, or retroactively delete the property in the
917 configuration with:
919 ----
920 qm set VMID -delete vmgenid
921 ----
923 The most prominent use case for 'vmgenid' are newer Microsoft Windows
924 operating systems, which use it to avoid problems in time sensitive or
925 replicate services (e.g., databases, domain controller
926 footnote:[])
927 on snapshot rollback, backup restore or a whole VM clone operation.
929 Importing Virtual Machines and disk images
930 ------------------------------------------
932 A VM export from a foreign hypervisor takes usually the form of one or more disk
933 images, with a configuration file describing the settings of the VM (RAM,
934 number of cores). +
935 The disk images can be in the vmdk format, if the disks come from
936 VMware or VirtualBox, or qcow2 if the disks come from a KVM hypervisor.
937 The most popular configuration format for VM exports is the OVF standard, but in
938 practice interoperation is limited because many settings are not implemented in
939 the standard itself, and hypervisors export the supplementary information
940 in non-standard extensions.
942 Besides the problem of format, importing disk images from other hypervisors
943 may fail if the emulated hardware changes too much from one hypervisor to
944 another. Windows VMs are particularly concerned by this, as the OS is very
945 picky about any changes of hardware. This problem may be solved by
946 installing the utility available from the Internet before exporting
947 and choosing a hard disk type of *IDE* before booting the imported Windows VM.
949 Finally there is the question of paravirtualized drivers, which improve the
950 speed of the emulated system and are specific to the hypervisor.
951 GNU/Linux and other free Unix OSes have all the necessary drivers installed by
952 default and you can switch to the paravirtualized drivers right after importing
953 the VM. For Windows VMs, you need to install the Windows paravirtualized
954 drivers by yourself.
956 GNU/Linux and other free Unix can usually be imported without hassle. Note
957 that we cannot guarantee a successful import/export of Windows VMs in all
958 cases due to the problems above.
960 Step-by-step example of a Windows OVF import
961 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
963 Microsoft provides
964[Virtual Machines downloads]
965 to get started with Windows development.We are going to use one of these
966 to demonstrate the OVF import feature.
968 Download the Virtual Machine zip
969 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
971 After getting informed about the user agreement, choose the _Windows 10
972 Enterprise (Evaluation - Build)_ for the VMware platform, and download the zip.
974 Extract the disk image from the zip
975 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
977 Using the `unzip` utility or any archiver of your choice, unpack the zip,
978 and copy via ssh/scp the ovf and vmdk files to your {pve} host.
980 Import the Virtual Machine
981 ^^^^^^^^^^^^^^^^^^^^^^^^^^
983 This will create a new virtual machine, using cores, memory and
984 VM name as read from the OVF manifest, and import the disks to the +local-lvm+
985 storage. You have to configure the network manually.
987 qm importovf 999 WinDev1709Eval.ovf local-lvm
989 The VM is ready to be started.
991 Adding an external disk image to a Virtual Machine
992 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
994 You can also add an existing disk image to a VM, either coming from a
995 foreign hypervisor, or one that you created yourself.
997 Suppose you created a Debian/Ubuntu disk image with the 'vmdebootstrap' tool:
999 vmdebootstrap --verbose \
1000 --size 10GiB --serial-console \
1001 --grub --no-extlinux \
1002 --package openssh-server \
1003 --package avahi-daemon \
1004 --package qemu-guest-agent \
1005 --hostname vm600 --enable-dhcp \
1006 --customize=./ \
1007 --sparse --image vm600.raw
1009 You can now create a new target VM for this image.
1011 qm create 600 --net0 virtio,bridge=vmbr0 --name vm600 --serial0 socket \
1012 --bootdisk scsi0 --scsihw virtio-scsi-pci --ostype l26
1014 Add the disk image as +unused0+ to the VM, using the storage +pvedir+:
1016 qm importdisk 600 vm600.raw pvedir
1018 Finally attach the unused disk to the SCSI controller of the VM:
1020 qm set 600 --scsi0 pvedir:600/vm-600-disk-1.raw
1022 The VM is ready to be started.
1025 ifndef::wiki[]
1026 include::qm-cloud-init.adoc[]
1027 endif::wiki[]
1029 ifndef::wiki[]
1030 include::qm-pci-passthrough.adoc[]
1031 endif::wiki[]
1034 Managing Virtual Machines with `qm`
1035 ------------------------------------
1037 qm is the tool to manage Qemu/Kvm virtual machines on {pve}. You can
1038 create and destroy virtual machines, and control execution
1039 (start/stop/suspend/resume). Besides that, you can use qm to set
1040 parameters in the associated config file. It is also possible to
1041 create and delete virtual disks.
1043 CLI Usage Examples
1044 ~~~~~~~~~~~~~~~~~~
1046 Using an iso file uploaded on the 'local' storage, create a VM
1047 with a 4 GB IDE disk on the 'local-lvm' storage
1049 qm create 300 -ide0 local-lvm:4 -net0 e1000 -cdrom local:iso/proxmox-mailgateway_2.1.iso
1051 Start the new VM
1053 qm start 300
1055 Send a shutdown request, then wait until the VM is stopped.
1057 qm shutdown 300 && qm wait 300
1059 Same as above, but only wait for 40 seconds.
1061 qm shutdown 300 && qm wait 300 -timeout 40
1064 [[qm_configuration]]
1065 Configuration
1066 -------------
1068 VM configuration files are stored inside the Proxmox cluster file
1069 system, and can be accessed at `/etc/pve/qemu-server/<VMID>.conf`.
1070 Like other files stored inside `/etc/pve/`, they get automatically
1071 replicated to all other cluster nodes.
1073 NOTE: VMIDs < 100 are reserved for internal purposes, and VMIDs need to be
1074 unique cluster wide.
1076 .Example VM Configuration
1077 ----
1078 cores: 1
1079 sockets: 1
1080 memory: 512
1081 name: webmail
1082 ostype: l26
1083 bootdisk: virtio0
1084 net0: e1000=EE:D2:28:5F:B6:3E,bridge=vmbr0
1085 virtio0: local:vm-100-disk-1,size=32G
1086 ----
1088 Those configuration files are simple text files, and you can edit them
1089 using a normal text editor (`vi`, `nano`, ...). This is sometimes
1090 useful to do small corrections, but keep in mind that you need to
1091 restart the VM to apply such changes.
1093 For that reason, it is usually better to use the `qm` command to
1094 generate and modify those files, or do the whole thing using the GUI.
1095 Our toolkit is smart enough to instantaneously apply most changes to
1096 running VM. This feature is called "hot plug", and there is no
1097 need to restart the VM in that case.
1100 File Format
1101 ~~~~~~~~~~~
1103 VM configuration files use a simple colon separated key/value
1104 format. Each line has the following format:
1106 -----
1107 # this is a comment
1108 OPTION: value
1109 -----
1111 Blank lines in those files are ignored, and lines starting with a `#`
1112 character are treated as comments and are also ignored.
1115 [[qm_snapshots]]
1116 Snapshots
1117 ~~~~~~~~~
1119 When you create a snapshot, `qm` stores the configuration at snapshot
1120 time into a separate snapshot section within the same configuration
1121 file. For example, after creating a snapshot called ``testsnapshot'',
1122 your configuration file will look like this:
1124 .VM configuration with snapshot
1125 ----
1126 memory: 512
1127 swap: 512
1128 parent: testsnaphot
1129 ...
1131 [testsnaphot]
1132 memory: 512
1133 swap: 512
1134 snaptime: 1457170803
1135 ...
1136 ----
1138 There are a few snapshot related properties like `parent` and
1139 `snaptime`. The `parent` property is used to store the parent/child
1140 relationship between snapshots. `snaptime` is the snapshot creation
1141 time stamp (Unix epoch).
1144 [[qm_options]]
1145 Options
1146 ~~~~~~~
1148 include::qm.conf.5-opts.adoc[]
1151 Locks
1152 -----
1154 Online migrations, snapshots and backups (`vzdump`) set a lock to
1155 prevent incompatible concurrent actions on the affected VMs. Sometimes
1156 you need to remove such a lock manually (e.g., after a power failure).
1158 qm unlock <vmid>
1160 CAUTION: Only do that if you are sure the action which set the lock is
1161 no longer running.
1164 ifdef::wiki[]
1166 See Also
1167 ~~~~~~~~
1169 * link:/wiki/Cloud-Init_Support[Cloud-Init Support]
1171 endif::wiki[]
1174 ifdef::manvolnum[]
1176 Files
1177 ------
1179 `/etc/pve/qemu-server/<VMID>.conf`::
1181 Configuration file for the VM '<VMID>'.
1184 include::pve-copyright.adoc[]
1185 endif::manvolnum[]