local-lvm.adoc: move Hardware into separate subsection, add info about bootloader
[pve-docs.git] / qm.adoc
CommitLineData
f69cfd23
DM
1ifdef::manvolnum[]
2PVE({manvolnum})
3================
38fd0958 4include::attributes.txt[]
f69cfd23
DM
5
6NAME
7----
8
9qm - Qemu/KVM Virtual Machine Manager
10
11
12SYNOPSYS
13--------
14
15include::qm.1-synopsis.adoc[]
16
17DESCRIPTION
18-----------
19endif::manvolnum[]
20
21ifndef::manvolnum[]
22Qemu/KVM Virtual Machines
23=========================
38fd0958 24include::attributes.txt[]
f69cfd23
DM
25endif::manvolnum[]
26
c4cba5d7
EK
27// deprecates
28// http://pve.proxmox.com/wiki/Container_and_Full_Virtualization
29// http://pve.proxmox.com/wiki/KVM
30// http://pve.proxmox.com/wiki/Qemu_Server
31
32Qemu (short form for Quick Emulator) is an opensource hypervisor that emulates a
33physical computer. From the perspective of the host system where Qemu is
34running, Qemu is a user program which has access to a number of local resources
35like partitions, files, network cards which are then passed to an
189d3661 36emulated computer which sees them as if they were real devices.
c4cba5d7
EK
37
38A guest operating system running in the emulated computer accesses these
39devices, and runs as it were running on real hardware. For instance you can pass
40an iso image as a parameter to Qemu, and the OS running in the emulated computer
189d3661 41will see a real CDROM inserted in a CD drive.
c4cba5d7 42
189d3661 43Qemu can emulates a great variety of hardware from ARM to Sparc, but {pve} is
c4cba5d7
EK
44only concerned with 32 and 64 bits PC clone emulation, since it represents the
45overwhelming majority of server hardware. The emulation of PC clones is also one
46of the fastest due to the availability of processor extensions which greatly
47speed up Qemu when the emulated architecture is the same as the host
9c63b5d9
EK
48architecture.
49
50NOTE: You may sometimes encounter the term _KVM_ (Kernel-based Virtual Machine).
51It means that Qemu is running with the support of the virtualization processor
52extensions, via the Linux kvm module. In the context of {pve} _Qemu_ and
53_KVM_ can be use interchangeably as Qemu in {pve} will always try to load the kvm
54module.
55
c4cba5d7
EK
56Qemu inside {pve} runs as a root process, since this is required to access block
57and PCI devices.
58
59Emulated devices and paravirtualized devices
60--------------------------------------------
61
189d3661
DC
62The PC hardware emulated by Qemu includes a mainboard, network controllers,
63scsi, ide and sata controllers, serial ports (the complete list can be seen in
64the `kvm(1)` man page) all of them emulated in software. All these devices
65are the exact software equivalent of existing hardware devices, and if the OS
66running in the guest has the proper drivers it will use the devices as if it
c4cba5d7
EK
67were running on real hardware. This allows Qemu to runs _unmodified_ operating
68systems.
69
70This however has a performance cost, as running in software what was meant to
71run in hardware involves a lot of extra work for the host CPU. To mitigate this,
72Qemu can present to the guest operating system _paravirtualized devices_, where
73the guest OS recognizes it is running inside Qemu and cooperates with the
74hypervisor.
75
76Qemu relies on the virtio virtualization standard, and is thus able to presente
189d3661
DC
77paravirtualized virtio devices, which includes a paravirtualized generic disk
78controller, a paravirtualized network card, a paravirtualized serial port,
c4cba5d7
EK
79a paravirtualized SCSI controller, etc ...
80
189d3661
DC
81It is highly recommended to use the virtio devices whenever you can, as they
82provide a big performance improvement. Using the virtio generic disk controller
83versus an emulated IDE controller will double the sequential write throughput,
84as measured with `bonnie++(8)`. Using the virtio network interface can deliver
c4cba5d7 85up to three times the throughput of an emulated Intel E1000 network card, as
189d3661 86measured with `iperf(1)`. footnote:[See this benchmark on the KVM wiki
c4cba5d7
EK
87http://www.linux-kvm.org/page/Using_VirtIO_NIC]
88
89Virtual Machines settings
90-------------------------
91Generally speaking {pve} tries to choose sane defaults for virtual machines
92(VM). Make sure you understand the meaning of the settings you change, as it
93could incur a performance slowdown, or putting your data at risk.
94
95General Settings
96~~~~~~~~~~~~~~~~
97General settings of a VM include
98
99* the *Node* : the physical server on which the VM will run
100* the *VM ID*: a unique number in this {pve} installation used to identify your VM
101* *Name*: a free form text string you can use to describe the VM
102* *Resource Pool*: a logical group of VMs
103
104OS Settings
105~~~~~~~~~~~
106When creating a VM, setting the proper Operating System(OS) allows {pve} to
107optimize some low level parameters. For instance Windows OS expect the BIOS
108clock to use the local time, while Unix based OS expect the BIOS clock to have
109the UTC time.
110
111Hard Disk
112~~~~~~~~~
2ec49380 113Qemu can emulate a number of storage controllers:
c4cba5d7
EK
114
115* the *IDE* controller, has a design which goes back to the 1984 PC/AT disk
116controller. Even if this controller has been superseded by more more designs,
117each and every OS you can think has support for it, making it a great choice
118if you want to run an OS released before 2003. You can connect up to 4 devices
119on this controller.
120
121* the *SATA* (Serial ATA) controller, dating from 2003, has a more modern
122design, allowing higher throughput and a greater number of devices to be
123connected. You can connect up to 6 devices on this controller.
124
125* the *SCSI* controller, designed in 1985, is commonly found on server
189d3661 126grade hardware, and can connect up to 14 storage devices. {pve} emulates by
c4cba5d7
EK
127default a LSI 53C895A controller.
128
129* The *Virtio* controller is a generic paravirtualized controller, and is the
130recommended setting if you aim for performance. To use this controller, the OS
131need to have special drivers which may be included in your installation ISO or
132not. Linux distributions have support for the Virtio controller since 2010, and
133FreeBSD since 2014. For Windows OSes, you need to provide an extra iso
189d3661 134containing the Virtio drivers during the installation.
c4cba5d7
EK
135// see: https://pve.proxmox.com/wiki/Paravirtualized_Block_Drivers_for_Windows#During_windows_installation.
136You can connect up to 16 devices on this controller.
137
138On each controller you attach a number of emulated hard disks, which are backed
139by a file or a block device residing in the configured storage. The choice of
140a storage type will determine the format of the hard disk image. Storages which
141present block devices (LVM, ZFS, Ceph) will require the *raw disk image format*,
142whereas files based storages (Ext4, NFS, GlusterFS) will let you to choose
143either the *raw disk image format* or the *QEMU image format*.
144
145 * the *QEMU image format* is a copy on write format which allows snapshots, and
146 thin provisioning of the disk image.
189d3661
DC
147 * the *raw disk image* is a bit-to-bit image of a hard disk, similar to what
148 you would get when executing the `dd` command on a block device in Linux. This
149 format do not support thin provisioning or snapshotting by itself, requiring
150 cooperation from the storage layer for these tasks. It is however 10% faster
151 than the *QEMU image format*. footnote:[See this benchmark for details
c4cba5d7 152 http://events.linuxfoundation.org/sites/events/files/slides/CloudOpen2013_Khoa_Huynh_v3.pdf]
189d3661 153 * the *VMware image format* only makes sense if you intend to import/export the
c4cba5d7
EK
154 disk image to other hypervisors.
155
156Setting the *Cache* mode of the hard drive will impact how the host system will
157notify the guest systems of block write completions. The *No cache* default
158means that the guest system will be notified that a write is complete when each
159block reaches the physical storage write queue, ignoring the host page cache.
160This provides a good balance between safety and speed.
161
162If you want the {pve} backup manager to skip a disk when doing a backup of a VM,
163you can set the *No backup* option on that disk.
164
165If your storage supports _thin provisioning_ (see the storage chapter in the
166{pve} guide), and your VM has a *SCSI* controller you can activate the *Discard*
167option on the hard disks connected to that controller. With *Discard* enabled,
168when the filesystem of a VM marks blocks as unused after removing files, the
169emulated SCSI controller will relay this information to the storage, which will
170then shrink the disk image accordingly.
171
af9c6de1 172.IO Thread
2b6e4b66
EK
173The option *IO Thread* can only be enabled when using a disk with the *VirtIO* controller,
174or with the *SCSI* controller, when the emulated controller type is *VirtIO SCSI*.
c564fc52
DC
175With this enabled, Qemu uses one thread per disk, instead of one thread for all,
176so it should increase performance when using multiple disks.
177Note that backups do not currently work with *IO Thread* enabled.
178
34e541c5
EK
179CPU
180~~~
181A *CPU socket* is a physical slot on a PC motherboard where you can plug a CPU.
182This CPU can then contain one or many *cores*, which are independent
183processing units. Whether you have a single CPU socket with 4 cores, or two CPU
184sockets with two cores is mostly irrelevant from a performance point of view.
185However some software is licensed depending on the number of sockets you have in
186your machine, in that case it makes sense to set the number of of sockets to
187what the license allows you, and increase the number of cores. +
188Increasing the number of virtual cpus (cores and sockets) will usually provide a
189performance improvement though that is heavily dependent on the use of the VM.
190Multithreaded applications will of course benefit from a large number of
191virtual cpus, as for each virtual cpu you add, Qemu will create a new thread of
192execution on the host system. If you're not sure about the workload of your VM,
193it is usually a safe bet to set the number of *Total cores* to 2.
194
195NOTE: It is perfectly safe to set the _overall_ number of total cores in all
196your VMs to be greater than the number of of cores you have on your server (ie.
1974 VMs with each 4 Total cores running in a 8 core machine is OK) In that case
198the host system will balance the Qemu execution threads between your server
199cores just like if you were running a standard multithreaded application.
200However {pve} will prevent you to allocate on a _single_ machine more vcpus than
201physically available, as this will only bring the performance down due to the
202cost of context switches.
203
204Qemu can emulate a number different of *CPU types* from 486 to the latest Xeon
205processors. Each new processor generation adds new features, like hardware
206assisted 3d rendering, random number generation, memory protection, etc ...
207Usually you should select for your VM a processor type which closely matches the
208CPU of the host system, as it means that the host CPU features (also called _CPU
209flags_ ) will be available in your VMs. If you want an exact match, you can set
210the CPU type to *host* in which case the VM will have exactly the same CPU flags
211as your host system. +
212This has a downside though. If you want to do a live migration of VMs between
213different hosts, your VM might end up on a new system with a different CPU type.
214If the CPU flags passed to the guest are missing, the qemu process will stop. To
215remedy this Qemu has also its own CPU type *kvm64*, that {pve} uses by defaults.
216kvm64 is a Pentium 4 look a like CPU type, which has a reduced CPU flags set,
217but is guaranteed to work everywhere. +
218 In short, if you care about live migration and moving VMs between nodes, leave
219the kvm64 default. If you don’t care about live migration, set the CPU type to
220host, as in theory this will give your guests maximum performance.
221
222You can also optionally emulate a *NUMA* architecture in your VMs. The basics of
223the NUMA architecture mean that instead of having a global memory pool available
224to all your cores, the memory is spread into local banks close to each socket.
225This can bring speed improvements as the memory bus is not a bottleneck
226anymore. If your system has a NUMA architecture footnote:[if the command
227`numactl --hardware | grep available` returns more than one node, then your host
228system has a NUMA architecture] we recommend to activate the option, as this
229will allow proper distribution of the VM resources on the host system. This
230option is also required in {pve} to allow hotplugging of cores and RAM to a VM.
231
232If the NUMA option is used, it is recommended to set the number of sockets to
233the number of sockets of the host system.
234
235Memory
236~~~~~~
237For each VM you have the option to set a fixed size memory or asking
238{pve} to dynamically allocate memory based on the current RAM usage of the
239host.
240
241When choosing a *fixed size memory* {pve} will simply allocate what you
242specify to your VM.
243
244// see autoballoon() in pvestatd.pm
245When choosing to *automatically allocate memory*, {pve} will make sure that the
246minimum amount you specified is always available to the VM, and if RAM usage on
247the host is below 80%, will dynamically add memory to the guest up to the
248maximum memory specified. +
249When the host is becoming short on RAM, the VM will then release some memory
250back to the host, swapping running processes if needed and starting the oom
251killer in last resort. The passing around of memory between host and guest is
252done via a special `balloon` kernel driver running inside the guest, which will
253grab or release memory pages from the host.
254footnote:[A good explanation of the inner workings of the balloon driver can be found here https://rwmj.wordpress.com/2010/07/17/virtio-balloon/]
255
c9f6e1a4
EK
256When multiple VMs use the autoallocate facility, it is possible to set a
257*Shares* coefficient which indicates the relative amount of the free host memory
258that each VM shoud take. Suppose for instance you have four VMs, three of them
259running a HTTP server and the last one is a database server. To cache more
260database blocks in the database server RAM, you would like to prioritize the
261database VM when spare RAM is available. For this you assign a Shares property
262of 3000 to the database VM, leaving the other VMs to the Shares default setting
263of 1000. The host server has 32GB of RAM, and is curring using 16GB, leaving 32
264* 80/100 - 16 = 9GB RAM to be allocated to the VMs. The database VM will get 9 *
2653000 / (3000 + 1000 + 1000 + 1000) = 4.5 GB extra RAM and each HTTP server will
266get 1/5 GB.
267
34e541c5
EK
268All Linux distributions released after 2010 have the balloon kernel driver
269included. For Windows OSes, the balloon driver needs to be added manually and can
270incur a slowdown of the guest, so we don't recommend using it on critical
271systems.
272// see https://forum.proxmox.com/threads/solved-hyper-threading-vs-no-hyper-threading-fixed-vs-variable-memory.20265/
273
274When allocating RAMs to your VMs, a good rule of thumb is always to leave 1GB
275of RAM available to the host.
276
1ff7835b
EK
277Network Device
278~~~~~~~~~~~~~~
279Each VM can have many _Network interface controllers_ (NIC), of four different
280types:
281
282 * *Intel E1000* is the default, and emulates an Intel Gigabit network card.
283 * the *VirtIO* paravirtualized NIC should be used if you aim for maximum
284performance. Like all VirtIO devices, the guest OS should have the proper driver
285installed.
286 * the *Realtek 8139* emulates an older 100 MB/s network card, and should
287only be used when emulating older operating systems ( released before 2002 )
288 * the *vmxnet3* is another paravirtualized device, which should only be used
289when importing a VM from another hypervisor.
290
291{pve} will generate for each NIC a random *MAC address*, so that your VM is
292addressable on Ethernet networks.
293
af9c6de1
EK
294The NIC you added to the VM can follow one of two differents models:
295
296 * in the default *Bridged mode* each virtual NIC is backed on the host by a
297_tap device_, ( a software loopback device simulating an Ethernet NIC ). This
298tap device is added to a bridge, by default vmbr0 in {pve}. In this mode, VMs
299have direct access to the Ethernet LAN on which the host is located.
300 * in the alternative *NAT mode*, each virtual NIC will only communicate with
301the Qemu user networking stack, where a builting router and DHCP server can
302provide network access. This built-in DHCP will serve adresses in the private
30310.0.2.0/24 range. The NAT mode is much slower than the bridged mode, and
304should only be used for testing.
305
306You can also skip adding a network device when creating a VM by selecting *No
307network device*.
308
309.Multiqueue
1ff7835b 310If you are using the VirtIO driver, you can optionally activate the
af9c6de1 311*Multiqueue* option. This option allows the guest OS to process networking
1ff7835b
EK
312packets using multiple virtual CPUs, providing an increase in the total number
313of packets transfered.
314
315//http://blog.vmsplice.net/2011/09/qemu-internals-vhost-architecture.html
316When using the VirtIO driver with {pve}, each NIC network queue is passed to the
317host kernel, where the queue will be processed by a kernel thread spawn by the
318vhost driver. With this option activated, it is possible to pass _multiple_
319network queues to the host kernel for each NIC.
320
321//https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/Virtualization_Tuning_and_Optimization_Guide/sect-Virtualization_Tuning_Optimization_Guide-Networking-Techniques.html#sect-Virtualization_Tuning_Optimization_Guide-Networking-Multi-queue_virtio-net
af9c6de1 322When using Multiqueue, it is recommended to set it to a value equal
1ff7835b
EK
323to the number of Total Cores of your guest. You also need to set in
324the VM the number of multi-purpose channels on each VirtIO NIC with the ethtool
325command:
326
327`ethtool -L eth0 combined X`
328
329where X is the number of the number of vcpus of the VM.
330
af9c6de1 331You should note that setting the Multiqueue parameter to a value greater
1ff7835b
EK
332than one will increase the CPU load on the host and guest systems as the
333traffic increases. We recommend to set this option only when the VM has to
334process a great number of incoming connections, such as when the VM is running
335as a router, reverse proxy or a busy HTTP server doing long polling.
336
685cc8e0
DC
337USB Passthrough
338~~~~~~~~~~~~~~~
339There are two different types of USB passthrough devices:
340
341* Host USB passtrough
342* SPICE USB passthrough
343
344Host USB passthrough works by giving a VM a USB device of the host.
345This can either be done via the vendor- and product-id, or
346via the host bus and port.
347
348The vendor/product-id looks like this: *0123:abcd*,
349where *0123* is the id of the vendor, and *abcd* is the id
350of the product, meaning two pieces of the same usb device
351have the same id.
352
353The bus/port looks like this: *1-2.3.4*, where *1* is the bus
354and *2.3.4* is the port path. This represents the physical
355ports of your host (depending of the internal order of the
356usb controllers).
357
358If a device is present in a VM configuration when the VM starts up,
359but the device is not present in the host, the VM can boot without problems.
360As soon as the device/port ist available in the host, it gets passed through.
361
362WARNING: Using this kind of USB passthrough, means that you cannot move
363a VM online to another host, since the hardware is only available
364on the host the VM is currently residing.
365
366The second type of passthrough is SPICE USB passthrough. This is useful
367if you use a SPICE client which supports it. If you add a SPICE USB port
368to your VM, you can passthrough a USB device from where your SPICE client is,
369directly to the VM (for example an input device or hardware dongle).
370
dd042288
EK
371Managing Virtual Machines with 'qm'
372------------------------------------
f69cfd23 373
dd042288 374qm is the tool to manage Qemu/Kvm virtual machines on {pve}. You can
f69cfd23
DM
375create and destroy virtual machines, and control execution
376(start/stop/suspend/resume). Besides that, you can use qm to set
377parameters in the associated config file. It is also possible to
378create and delete virtual disks.
379
dd042288
EK
380CLI Usage Examples
381~~~~~~~~~~~~~~~~~~
382
383Create a new VM with 4 GB IDE disk.
384
385 qm create 300 -ide0 4 -net0 e1000 -cdrom proxmox-mailgateway_2.1.iso
386
387Start the new VM
388
389 qm start 300
390
391Send a shutdown request, then wait until the VM is stopped.
392
393 qm shutdown 300 && qm wait 300
394
395Same as above, but only wait for 40 seconds.
396
397 qm shutdown 300 && qm wait 300 -timeout 40
398
f69cfd23
DM
399Configuration
400-------------
401
402All configuration files consists of lines in the form
403
404 PARAMETER: value
405
871e1fd6 406Configuration files are stored inside the Proxmox cluster file
c4cba5d7 407system, and can be accessed at '/etc/pve/qemu-server/<VMID>.conf'.
f69cfd23 408
a7f36905
DM
409Options
410~~~~~~~
411
412include::qm.conf.5-opts.adoc[]
413
f69cfd23
DM
414
415Locks
416-----
417
871e1fd6
FG
418Online migrations and backups ('vzdump') set a lock to prevent incompatible
419concurrent actions on the affected VMs. Sometimes you need to remove such a
420lock manually (e.g., after a power failure).
f69cfd23
DM
421
422 qm unlock <vmid>
423
f69cfd23
DM
424
425ifdef::manvolnum[]
426include::pve-copyright.adoc[]
427endif::manvolnum[]