1virt-v2v(1) Virtualization Support virt-v2v(1)
2
6 virt-v2v - Convert a guest to use KVM
7
9 virt-v2v [-i mode] [other -i* options]
10 [-o mode] [other -o* options]
11 [guest|filename]
12 virt-v2v --in-place
14 [-i mode] [other -i* options]
15 [guest|filename]
16
18 Virt-v2v converts a single guest from a foreign hypervisor to run on
19 KVM. It can read Linux and Windows guests running on VMware, Xen,
20 Hyper-V and some other hypervisors, and convert them to KVM managed by
21 libvirt, OpenStack, oVirt, Red Hat Virtualisation (RHV) or several
22 other targets. It can modify the guest to make it bootable on KVM and
23 install virtio drivers so it will run quickly.
24 There is also a companion front-end called virt-p2v(1) which comes as
26 an ISO, CD or PXE image that can be booted on physical machines to
27 virtualize those machines (physical to virtual, or p2v).
28 Input and Output
30 You normally run virt-v2v with several -i* options controlling the
31 input mode and also several -o* options controlling the output mode.
32 In this sense, "input" refers to the source foreign hypervisor such as
33 VMware, and "output" refers to the target KVM-based management system
34 such as oVirt or OpenStack.
35 The input and output sides of virt-v2v are separate and unrelated.
37 Virt-v2v can read from any input and write to any output. Therefore
38 these sides of virt-v2v are documented separately in this manual.
39 Virt-v2v normally copies from the input to the output, called "copying
41 mode". In this case the source guest is always left unchanged. In-
42 place conversion (--in-place) only uses the -i* options and modifies
43 the source guest in-place. (See "In-place conversion" below.)
44 Other virt-v2v topics
46 virt-v2v-support(1) — Supported hypervisors, virtualization management
47 systems, guests.
48 virt-v2v-input-vmware(1) — Input from VMware.
50 virt-v2v-input-xen(1) — Input from Xen.
52 virt-v2v-output-local(1) — Output to local files or local libvirt.
54 virt-v2v-output-rhv(1) — Output to oVirt or RHV.
56 virt-v2v-output-openstack(1) — Output to OpenStack.
58 virt-v2v-copy-to-local(1) — Deprecated tool to handle Xen guests using
60 host block device storage.
61
63 Convert from VMware vCenter server to local libvirt
64 You have a VMware vCenter server called "vcenter.example.com", a
65 datacenter called "Datacenter", and an ESXi hypervisor called "esxi".
66 You want to convert a guest called "vmware_guest" to run locally under
67 libvirt.
68 virt-v2v -ic vpx://vcenter.example.com/Datacenter/esxi vmware_guest
70 In this case you will most likely have to run virt-v2v as "root", since
72 it needs to talk to the system libvirt daemon and copy the guest disks
73 to /var/lib/libvirt/images.
74 For more information see virt-v2v-input-vmware(1).
76 Convert from VMware to RHV/oVirt
78 This is the same as the previous example, except you want to send the
79 guest to a RHV Data Domain using the RHV REST API. Guest network
80 interface(s) are connected to the target network called "ovirtmgmt".
81 virt-v2v -ic vpx://vcenter.example.com/Datacenter/esxi vmware_guest \
83 -o rhv-upload -oc https://ovirt-engine.example.com/ovirt-engine/api \
84 -os ovirt-data -op /tmp/ovirt-admin-password -of raw \
85 -oo rhv-cafile=/tmp/ca.pem -oo rhv-direct \
86 --bridge ovirtmgmt
87 In this case the host running virt-v2v acts as a conversion server.
89 For more information see virt-v2v-output-rhv(1).
91 Convert from ESXi hypervisor over SSH to local libvirt
93 You have an ESXi hypervisor called "esxi.example.com" with SSH access
94 enabled. You want to convert from VMFS storage on that server to a
95 local file.
96 virt-v2v \
98 -i vmx -it ssh \
99 "ssh://root@esxi.example.com/vmfs/volumes/datastore1/guest/guest.vmx" \
100 -o local -os /var/tmp
101 The guest must not be running. Virt-v2v would not need to be run as
103 root in this case.
104 For more information about converting from VMX files see
106 virt-v2v-input-vmware(1).
107 Convert disk image to OpenStack
109 Given a disk image from another hypervisor that you want to convert to
110 run on OpenStack (only KVM-based OpenStack is supported), you can run
111 virt-v2v inside an OpenStack VM (called "v2v-vm" below), and do:
112 virt-v2v -i disk disk.img -o openstack -oo server-id=v2v-vm
114 See virt-v2v-output-openstack(1).
116 Convert disk image to disk image
118 Given a disk image from another hypervisor that you want to convert to
119 run on KVM, you have two options. The simplest way is to try:
120 virt-v2v -i disk disk.img -o local -os /var/tmp
122 where virt-v2v guesses everything about the input disk.img and (in this
124 case) writes the converted result to /var/tmp.
125 A more complex method is to write some libvirt XML describing the input
127 guest (if you can get the source hypervisor to provide you with libvirt
128 XML, then so much the better). You can then do:
129 virt-v2v -i libvirtxml guest-domain.xml -o local -os /var/tmp
131 Since guest-domain.xml contains the path(s) to the guest disk image(s)
133 you do not need to specify the name of the disk image on the command
134 line.
135 To convert a local disk image and immediately boot it in local qemu,
137 do:
138 virt-v2v -i disk disk.img -o qemu -os /var/tmp --qemu-boot
140
142 --help
143 Display help.
144 -b ...
146 --bridge ...
147 See --network below.
148 --colors
150 --colours
151 Use ANSI colour sequences to colourize messages. This is the
152 default when the output is a tty. If the output of the program is
153 redirected to a file, ANSI colour sequences are disabled unless you
154 use this option.
155 --compressed
157 Write a compressed output file. This is only allowed if the output
158 format is qcow2 (see -of below), and is equivalent to the -c option
159 of qemu-img(1).
160 --debug-overlays
162 Save the overlay file(s) created during conversion. This option is
163 only used for debugging virt-v2v and may be removed in a future
164 version.
165 --echo-keys
167 When prompting for keys and passphrases, virt-v2v normally turns
168 echoing off so you cannot see what you are typing. If you are not
169 worried about Tempest attacks and there is no one else in the room
170 you can specify this flag to see what you are typing.
171 Note this options only applies to keys and passphrases for
173 encrypted devices and partitions, not for passwords used to connect
174 to remote servers.
175 -i disk
177 Set the input method to disk.
178 In this mode you can read a virtual machine disk image with no
180 metadata. virt-v2v tries to guess the best default metadata. This
181 is usually adequate but you can get finer control (eg. of memory
182 and vCPUs) by using -i libvirtxml instead. Only guests that use a
183 single disk can be imported this way.
184 -i libvirt
186 Set the input method to libvirt. This is the default.
187 In this mode you have to specify a libvirt guest name or UUID on
189 the command line. You may also specify a libvirt connection URI
190 (see -ic).
191 -i libvirtxml
193 Set the input method to libvirtxml.
194 In this mode you have to pass a libvirt XML file on the command
196 line. This file is read in order to get metadata about the source
197 guest (such as its name, amount of memory), and also to locate the
198 input disks. See "Minimal XML for -i libvirtxml option" below.
199 -i local
201 This is the same as -i disk.
202 -i ova
204 Set the input method to ova.
205 In this mode you can read a VMware ova file. Virt-v2v will read
207 the ova manifest file and check the vmdk volumes for validity
208 (checksums) as well as analyzing the ovf file, and then convert the
209 guest. See virt-v2v-input-vmware(1).
210 -i vmx
212 Set the input method to vmx.
213 In this mode you can read a VMware vmx file directly or over SSH.
215 This is useful when VMware VMs are stored on an NFS server which
216 you can mount directly, or where you have access by SSH to an ESXi
217 hypervisor. See virt-v2v-input-vmware(1).
218 -ic libvirtURI
220 Specify a libvirt connection URI to use when reading the guest.
221 This is only used when -i libvirt.
222 Only local libvirt connections, VMware vCenter connections, or RHEL
224 5 Xen remote connections can be used. Other remote libvirt
225 connections will not work in general.
226 See also virt-v2v-input-vmware(1), virt-v2v-input-xen(1).
228 -if format
230 For -i disk only, this specifies the format of the input disk
231 image. For other input methods you should specify the input format
232 in the metadata.
233 --in-place
235 Do not create an output virtual machine in the target hypervisor.
236 Instead, adjust the guest OS in the source VM to run in the input
237 hypervisor.
238 This mode is meant for integration with other toolsets, which take
240 the responsibility of converting the VM configuration, providing
241 for rollback in case of errors, transforming the storage, etc.
242 See "In-place conversion" below.
244 Conflicts with all -o * options.
246 -io OPTION=VALUE
248 Set input option(s) related to the current input mode or transport.
249 To display short help on what options are available you can use:
250 virt-v2v -it vddk -io "?"
252 -io vddk-libdir=LIBDIR
254 Set the VDDK library directory. This directory should contain
255 subdirectories called include, lib64 etc., but do not include lib64
256 actually in the parameter.
257 In most cases this parameter is required when using the -it vddk
259 (VDDK) transport. See virt-v2v-input-vmware(1) for details.
260 -io vddk-thumbprint=xx:xx:xx:...
262 Set the thumbprint of the remote VMware server.
263 This parameter is required when using the -it vddk (VDDK)
265 transport. See virt-v2v-input-vmware(1) for details.
266 -io vddk-config=FILENAME
268 -io vddk-cookie=COOKIE
269 -io vddk-nfchostport=PORT
270 -io vddk-port=PORT
271 -io vddk-snapshot=SNAPSHOT-MOREF
272 -io vddk-transports=MODE:MODE:...
273 When using VDDK mode, these options are passed unmodified to the
274 nbdkit(1) VDDK plugin. Please refer to nbdkit-vddk-plugin(1). Do
275 not use these options unless you know what you are doing. These
276 are all optional.
277 -ip filename
279 Supply a file containing a password to be used when connecting to
280 the target hypervisor. If this is omitted then the input
281 hypervisor may ask for the password interactively. Note the file
282 should contain the whole password, without any trailing newline,
283 and for security the file should have mode 0600 so that others
284 cannot read it.
285 -it ssh
287 When using -i vmx, this enables the ssh transport. See
288 virt-v2v-input-vmware(1).
289 -it vddk
291 Use VMware VDDK as a transport to copy the input disks. See
292 virt-v2v-input-vmware(1). If you use this parameter then you may
293 need to use other -io vddk* options to specify how to connect
294 through VDDK.
295 --key SELECTOR
297 Specify a key for LUKS, to automatically open a LUKS device when
298 using the inspection. "SELECTOR" can be in one of the following
299 formats:
300 --key "DEVICE":key:KEY_STRING
302 Use the specified "KEY_STRING" as passphrase.
303 --key "DEVICE":file:FILENAME
305 Read the passphrase from FILENAME.
306 --keys-from-stdin
308 Read key or passphrase parameters from stdin. The default is to
309 try to read passphrases from the user by opening /dev/tty.
310 Note this options only applies to keys and passphrases for
312 encrypted devices and partitions, not for passwords used to connect
313 to remote servers.
314 --mac aa:bb:cc:dd:ee:ff:network:out
316 --mac aa:bb:cc:dd:ee:ff:bridge:out
317 Map source NIC MAC address to a network or bridge.
318 See "Networks and bridges" below.
320 --machine-readable
322 --machine-readable=format
323 This option is used to make the output more machine friendly when
324 being parsed by other programs. See "Machine readable output"
325 below.
326 -n in:out
328 -n out
329 --network in:out
330 --network out
331 -b in:out
332 -b out
333 --bridge in:out
334 --bridge out
335 Map network (or bridge) called "in" to network (or bridge) called
336 "out". If no "in:" prefix is given, all other networks (or
337 bridges) are mapped to "out".
338 See "Networks and bridges" below.
340 --no-copy
342 Don’t copy the disks. Instead, conversion is performed (and thrown
343 away), and metadata is written, but no disks are created. See also
344 discussion of -o null below.
345 This is useful in two cases: Either you want to test if conversion
347 is likely to succeed, without the long copying process. Or you are
348 only interested in looking at the metadata.
349 This option is not compatible with -o libvirt since it would create
351 a faulty guest (one with no disks).
352 This option is not compatible with -o glance for technical reasons.
354 -o disk
356 This is the same as -o local.
357 -o glance
359 This is a legacy option. You should probably use -o openstack
360 instead.
361 Set the output method to OpenStack Glance. In this mode the
363 converted guest is uploaded to Glance. See
364 virt-v2v-output-openstack(1).
365 -o libvirt
367 Set the output method to libvirt. This is the default.
368 In this mode, the converted guest is created as a libvirt guest.
370 You may also specify a libvirt connection URI (see -oc).
371 See virt-v2v-output-local(1).
373 -o local
375 Set the output method to local.
376 In this mode, the converted guest is written to a local directory
378 specified by -os /dir (the directory must exist). The converted
379 guest’s disks are written as:
380 /dir/name-sda
382 /dir/name-sdb
383 [etc]
384 and a libvirt XML file is created containing guest metadata:
386 /dir/name.xml
388 where "name" is the guest name.
390 -o null
392 Set the output method to null.
393 The guest is converted and copied (unless you also specify
395 --no-copy), but the results are thrown away and no metadata is
396 written.
397 -o openstack
399 Set the output method to OpenStack. See
400 virt-v2v-output-openstack(1).
401 -o ovirt
403 This is the same as -o rhv.
404 -o ovirt-upload
406 This is the same as -o rhv-upload.
407 -o qemu
409 Set the output method to qemu.
410 This is similar to -o local, except that a shell script is written
412 which you can use to boot the guest in qemu. The converted disks
413 and shell script are written to the directory specified by -os.
414 When using this output mode, you can also specify the --qemu-boot
416 option which boots the guest under qemu immediately.
417 -o rhev
419 This is the same as -o rhv.
420 -o rhv
422 Set the output method to rhv.
423 The converted guest is written to a RHV Export Storage Domain. The
425 -os parameter must also be used to specify the location of the
426 Export Storage Domain. Note this does not actually import the
427 guest into RHV. You have to do that manually later using the UI.
428 See virt-v2v-output-rhv(1).
430 -o rhv-upload
432 Set the output method to rhv-upload.
433 The converted guest is written directly to a RHV Data Domain. This
435 is a faster method than -o rhv, but requires oVirt or RHV ≥ 4.2.
436 See virt-v2v-output-rhv(1).
438 -o vdsm
440 Set the output method to vdsm.
441 This mode is similar to -o rhv, but the full path to the data
443 domain must be given:
444 /rhv/data-center/<data-center-uuid>/<data-domain-uuid>. This mode
445 is only used when virt-v2v runs under VDSM control.
446 -oa sparse
448 -oa preallocated
449 Set the output file allocation mode. The default is "sparse".
450 -oc URI
452 Specify a connection URI to use when writing the converted guest.
453 For -o libvirt this is the libvirt URI. Only local libvirt
455 connections can be used. Remote libvirt connections will not work.
456 See virt-v2v-output-local(1) for further information.
457 -of format
459 When converting the guest, convert the disks to the given format.
460 If not specified, then the input format is used.
462 -on name
464 Rename the guest when converting it. If this option is not used
465 then the output name is the same as the input name.
466 -oo OPTION=VALUE
468 Set output option(s) related to the current output mode. To
469 display short help on what options are available you can use:
470 virt-v2v -o rhv-upload -oo "?"
472 -oo guest-id="ID"
474 For -o openstack (virt-v2v-output-openstack(1)) only, set a guest
475 ID which is saved on each Cinder volume in the "virt_v2v_guest_id"
476 volume property.
477 -oo verify-server-certificate
479 -oo verify-server-certificate="true|false"
480 For -o openstack (virt-v2v-output-openstack(1)) only, this can be
481 used to disable SSL certification validation when connecting to
482 OpenStack by specifying -oo verify-server-certificate=false.
483 -oo os-*=*
485 For -o openstack (virt-v2v-output-openstack(1)) only, set optional
486 OpenStack authentication. For example -oo os-username=NAME is
487 equivalent to "openstack --os-username=NAME".
488 -oo rhv-cafile=ca.pem
490 For -o rhv-upload (virt-v2v-output-rhv(1)) only, the ca.pem file
491 (Certificate Authority), copied from /etc/pki/ovirt-engine/ca.pem
492 on the oVirt engine.
493 -oo rhv-cluster="CLUSTERNAME"
495 For -o rhv-upload (virt-v2v-output-rhv(1)) only, set the RHV
496 Cluster Name. If not given it uses "Default".
497 -oo rhv-direct
499 For -o rhv-upload (virt-v2v-output-rhv(1)) only, if this option is
500 given then virt-v2v will attempt to directly upload the disk to the
501 oVirt node, otherwise it will proxy the upload through the oVirt
502 engine. Direct upload requires that you have network access to the
503 oVirt nodes. Non-direct upload is slightly slower but should work
504 in all situations.
505 -oo rhv-verifypeer
507 For -o rhv-upload (virt-v2v-output-rhv(1)) only, verify the
508 oVirt/RHV server’s identity by checking the server‘s certificate
509 against the Certificate Authority.
510 -oo server-id="NAME|UUID"
512 For -o openstack (virt-v2v-output-openstack(1)) only, set the name
513 of the conversion appliance where virt-v2v is running.
514 -oo vdsm-compat=0.10
516 -oo vdsm-compat=1.1
517 If -o vdsm and the output format is qcow2, then we add the qcow2
518 compat=0.10 option to the output file for compatibility with RHEL 6
519 (see https://bugzilla.redhat.com/1145582).
520 If -oo vdsm-compat=1.1 is used then modern qcow2 (compat=1.1) files
522 are generated instead.
523 Currently -oo vdsm-compat=0.10 is the default, but this will change
525 to -oo vdsm-compat=1.1 in a future version of virt-v2v (when we can
526 assume that everyone is using a modern version of qemu).
527 Note this option only affects -o vdsm output. All other output
529 modes (including -o rhv) generate modern qcow2 compat=1.1 files,
530 always.
531 If this option is available, then "vdsm-compat-option" will appear
533 in the --machine-readable output.
534 -oo vdsm-image-uuid=UUID
536 -oo vdsm-vol-uuid=UUID
537 -oo vdsm-vm-uuid=UUID
538 -oo vdsm-ovf-output=DIR
539 Normally the RHV output mode chooses random UUIDs for the target
540 guest. However VDSM needs to control the UUIDs and passes these
541 parameters when virt-v2v runs under VDSM control. The parameters
542 control:
543 · the image directory of each guest disk (-oo vdsm-image-uuid)
545 (this option is passed once for each guest disk)
546 · UUIDs for each guest disk (-oo vdsm-vol-uuid) (this option is
548 passed once for each guest disk)
549 · the OVF file name (-oo vdsm-vm-uuid).
551 · the OVF output directory (default current directory) (-oo vdsm-
553 ovf-output).
554 The format of UUIDs is: "12345678-1234-1234-1234-123456789abc"
556 (each hex digit can be "0-9" or "a-f"), conforming to OSF DCE 1.1.
557 These options can only be used with -o vdsm.
559 -oo vdsm-ovf-flavour=flavour
561 This option controls the format of the OVF generated at the end of
562 conversion. Currently there are two possible flavours:
563 rhvexp
565 The OVF format used in RHV export storage domain.
566 ovirt
568 The OVF format understood by oVirt REST API.
569 For backward compatibility the default is rhvexp, but this may
571 change in the future.
572 -op file
574 Supply a file containing a password to be used when connecting to
575 the target hypervisor. Note the file should contain the whole
576 password, without any trailing newline, and for security the file
577 should have mode 0600 so that others cannot read it.
578 -os storage
580 The location of the storage for the converted guest.
581 For -o libvirt, this is a libvirt directory pool (see
583 "virsh pool-list") or pool UUID.
584 For -o local and -o qemu, this is a directory name. The directory
586 must exist.
587 For -o rhv-upload, this is the name of the destination Storage
589 Domain.
590 For -o openstack, this is the optional Cinder volume type.
592 For -o rhv, this can be an NFS path of the Export Storage Domain of
594 the form "<host>:<path>", eg:
595 rhv-storage.example.com:/rhv/export
597 The NFS export must be mountable and writable by the user and host
599 running virt-v2v, since the virt-v2v program has to actually mount
600 it when it runs. So you probably have to run virt-v2v as "root".
601 Or: You can mount the Export Storage Domain yourself, and point -os
603 to the mountpoint. Note that virt-v2v will still need to write to
604 this remote directory, so virt-v2v will still need to run as
605 "root".
606 You will get an error if virt-v2v is unable to mount/write to the
608 Export Storage Domain.
609 --print-estimate
611 Print the estimated size of the data which will be copied from the
612 source disk(s) and stop. One number (the size in bytes) is printed
613 per disk, and a total:
614 $ virt-v2v --print-estimate
616 ...
617 disk 1: 100000
618 disk 2: 200000
619 total: 300000
620 With the --machine-readable option you get JSON output which can be
622 directed into a file or elsewhere:
623 $ virt-v2v --print-estimate --machine-readable=file:estimates
625 ...
626 $ cat estimates
627 {
628 "disks": [ 100000, 200000 ],
629 "total": 300000
630 }
631 When using this option you must specify an output mode. This is
633 because virt-v2v has to perform the conversion in order to print
634 the estimate, and the conversion depends on the output mode. Using
635 -o null should be safe for most purposes.
636 When this option is used along with --machine-readable you can
638 direct the output to an alternate file.
639 --print-source
641 Print information about the source guest and stop. This option is
642 useful when you are setting up network and bridge maps. See
643 "Networks and bridges".
644 --qemu-boot
646 When using -o qemu only, this boots the guest immediately after
647 virt-v2v finishes.
648 -q
650 --quiet
651 This disables progress bars and other unnecessary output.
652 --root ask
654 --root single
655 --root first
656 --root /dev/sdX
657 --root /dev/VG/LV
658 Choose the root filesystem to be converted.
659 In the case where the virtual machine is dual-boot or multi-boot,
661 or where the VM has other filesystems that look like operating
662 systems, this option can be used to select the root filesystem
663 (a.k.a. "C:" drive or /) of the operating system that is to be
664 converted. The Windows Recovery Console, certain attached DVD
665 drives, and bugs in libguestfs inspection heuristics, can make a
666 guest look like a multi-boot operating system.
667 The default in virt-v2v ≤ 0.7.1 was --root single, which causes
669 virt-v2v to die if a multi-boot operating system is found.
670 Since virt-v2v ≥ 0.7.2 the default is now --root ask: If the VM is
672 found to be multi-boot, then virt-v2v will stop and list the
673 possible root filesystems and ask the user which to use. This
674 requires that virt-v2v is run interactively.
675 --root first means to choose the first root device in the case of a
677 multi-boot operating system. Since this is a heuristic, it may
678 sometimes choose the wrong one.
679 You can also name a specific root device, eg. --root /dev/sda2
681 would mean to use the second partition on the first hard drive. If
682 the named root device does not exist or was not detected as a root
683 device, then virt-v2v will fail.
684 Note that there is a bug in grub which prevents it from
686 successfully booting a multiboot system if virtio is enabled. Grub
687 is only able to boot an operating system from the first virtio
688 disk. Specifically, /boot must be on the first virtio disk, and it
689 cannot chainload an OS which is not in the first virtio disk.
690 -v
692 --verbose
693 Enable verbose messages for debugging.
694 -V
696 --version
697 Display version number and exit.
698 -x Enable tracing of libguestfs API calls.
700
702 Xen paravirtualized guests
703 Older versions of virt-v2v could turn a Xen paravirtualized (PV) guest
704 into a KVM guest by installing a new kernel. This version of virt-v2v
705 does not attempt to install any new kernels. Instead it will give you
706 an error if there are only Xen PV kernels available.
707 Therefore before conversion you should check that a regular kernel is
709 installed. For some older Linux distributions, this means installing a
710 kernel from the table below:
711 RHEL 3 (Does not apply, as there was no Xen PV kernel)
713 RHEL 4 i686 with > 10GB of RAM: install 'kernel-hugemem'
715 i686 SMP: install 'kernel-smp'
716 other i686: install 'kernel'
717 x86-64 SMP with > 8 CPUs: install 'kernel-largesmp'
718 x86-64 SMP: install 'kernel-smp'
719 other x86-64: install 'kernel'
720 RHEL 5 i686: install 'kernel-PAE'
722 x86-64: install 'kernel'
723 SLES 10 i586 with > 10GB of RAM: install 'kernel-bigsmp'
725 i586 SMP: install 'kernel-smp'
726 other i586: install 'kernel-default'
727 x86-64 SMP: install 'kernel-smp'
728 other x86-64: install 'kernel-default'
729 SLES 11+ i586: install 'kernel-pae'
731 x86-64: install 'kernel-default'
732 Windows (Does not apply, as there is no Xen PV Windows kernel)
734 Enabling virtio
736 "Virtio" is the name for a set of drivers which make disk (block
737 device), network and other guest operations work much faster on KVM.
738 Older versions of virt-v2v could install these drivers for certain
740 Linux guests. This version of virt-v2v does not attempt to install new
741 Linux kernels or drivers, but will warn you if they are not installed
742 already.
743 In order to enable virtio, and hence improve performance of the guest
745 after conversion, you should ensure that the minimum versions of
746 packages are installed before conversion, by consulting the table
747 below.
748 RHEL 3 No virtio drivers are available
750 RHEL 4 kernel >= 2.5.9-89.EL
752 lvm2 >= 2.02.42-5.el4
753 device-mapper >= 1.02.28-2.el4
754 selinux-policy-targeted >= 1.17.30-2.152.el4
755 policycoreutils >= 1.18.1-4.13
756 RHEL 5 kernel >= 2.6.18-128.el5
758 lvm2 >= 2.02.40-6.el5
759 selinux-policy-targeted >= 2.4.6-203.el5
760 RHEL 6+ All versions support virtio
762 Fedora All versions support virtio
764 SLES 11+ All versions support virtio
766 SLES 10 kernel >= 2.6.16.60-0.85.1
768 OpenSUSE 11+ All versions support virtio
770 OpenSUSE 10 kernel >= 2.6.25.5-1.1
772 Debian 6+ All versions support virtio
774 Ubuntu 10.04+ All versions support virtio
776 Windows Drivers are installed from the ISO or directory pointed
778 to by "VIRTIO_WIN" environment variable if present
779 RHEL 4: SELinux relabel appears to hang forever
781 In RHEL ≤ 4.7 there was a bug which causes SELinux relabelling to
782 appear to hang forever at:
783 *** Warning -- SELinux relabel is required. ***
785 *** Disabling security enforcement. ***
786 *** Relabeling could take a very long time, ***
787 *** depending on file system size. ***
788 In reality it is waiting for you to press a key (but there is no visual
790 indication of this). You can either hit the "[Return]" key, at which
791 point the guest will finish relabelling and reboot, or you can install
792 policycoreutils ≥ 1.18.1-4.13 before starting the v2v conversion. See
793 also https://bugzilla.redhat.com/show_bug.cgi?id=244636
794 Debian and Ubuntu
796 "warning: could not determine a way to update the configuration of
797 Grub2"
798 Currently, virt-v2v has no way to set the default kernel in Debian and
800 Ubuntu guests using GRUB 2 as bootloader. This means that virt-v2v
801 will not change the default kernel used for booting, even in case it is
802 not the best kernel available on the guest. A recommended procedure
803 is, before using virt-v2v, to check that the boot kernel is the best
804 kernel available in the guest (for example by making sure the guest is
805 up-to-date).
806 "vsyscall attempted with vsyscall=none"
808 When run on a recent Debian host virt-v2v may fail to convert guests
810 which were created before 2013. In the debugging output you will see a
811 crash message similar to:
812 vsyscall attempted with vsyscall=none ip:...
814 segfault at ...
815 This is caused because Debian removed support for running old binaries
817 which used the legacy vsyscall page to call into the kernel.
818 You can work around this problem by running this command before running
820 virt-v2v:
821 export LIBGUESTFS_APPEND="vsyscall=emulate"
823 For more information, see https://bugzilla.redhat.com/1592061
825 Windows
827 Windows ≥ 8 Fast Startup is incompatible with virt-v2v
828 Guests which use the Windows ≥ 8 "Fast Startup" feature (or guests
830 which are hibernated) cannot be converted with virt-v2v. You will see
831 an error:
832 virt-v2v: error: unable to mount the disk image for writing. This has
834 probably happened because Windows Hibernation or Fast Restart is being
835 used in this guest. You have to disable this (in the guest) in order
836 to use virt-v2v.
837 As the message says, you need to boot the guest and disable the "Fast
839 Startup" feature (Control Panel → Power Options → Choose what the power
840 buttons do → Change settings that are currently unavailable → Turn on
841 fast startup), and shut down the guest, and then you will be able to
842 convert it.
843 For more information, see: "WINDOWS HIBERNATION AND WINDOWS 8 FAST
845 STARTUP" in guestfs(3).
846 Boot failure: 0x0000007B
848 This boot failure is caused by Windows being unable to find or load the
850 right disk driver (eg. viostor.sys). If you experience this error,
851 here are some things to check:
852 · First ensure that the guest boots on the source hypervisor before
854 conversion.
855 · Check you have the Windows virtio drivers available in
857 /usr/share/virtio-win, and that virt-v2v did not print any warning
858 about not being able to install virtio drivers.
859 On Red Hat Enterprise Linux 7, you will need to install the signed
861 drivers available in the "virtio-win" package. If you do not have
862 access to the signed drivers, then you will probably need to
863 disable driver signing in the boot menus.
864 · Check that you are presenting a virtio-blk interface (not virtio-
866 scsi and not ide) to the guest. On the qemu/KVM command line you
867 should see something similar to this:
868 ... -drive file=windows-sda,if=virtio ...
870 In libvirt XML, you should see:
872 <target dev='vda' bus='virtio'/>
874 · Check that Windows Group Policy does not prevent the driver from
876 being installed or used. Try deleting Windows Group Policy before
877 conversion.
878 · Check there is no anti-virus or other software which implements
880 Group Policy-like prohibitions on installing or using new drivers.
881 · Enable boot debugging and check the viostor.sys driver is being
883 loaded.
884 OpenStack and Windows reactivation
886 OpenStack does not offer stable device / PCI addresses to guests.
888 Every time it creates or starts a guest, it regenerates the libvirt XML
889 for that guest from scratch. The libvirt XML will have no <address>
890 fields. Libvirt will then assign addresses to devices, in a
891 predictable manner. Addresses may change if any of the following are
892 true:
893 · A new disk or network device has been added or removed from the
895 guest.
896 · The version of OpenStack or (possibly) libvirt has changed.
898 Because Windows does not like "hardware" changes of this kind, it may
900 trigger Windows reactivation.
901 This can also prevent booting with a 7B error [see previous section] if
903 the guest has group policy containing "Device Installation
904 Restrictions".
905 Support for SHA-2 certificates in Windows 7 and Windows Server 2008 R2
907 Later versions of the Windows virtio drivers are signed using SHA-2
909 certificates (instead of SHA-1). The original shipping Windows 7 and
910 Windows Server 2008 R2 did not understand SHA-2 certificates and so the
911 Windows virtio drivers will not install properly.
912 To fix this you must apply SHA-2 Code Signing Support from:
914 https://docs.microsoft.com/en-us/security-updates/SecurityAdvisories/2015/3033929
915 before converting the guest.
916 For further information see:
918 https://bugzilla.redhat.com/show_bug.cgi?id=1624878
919 Networks and bridges
921 Guests are usually connected to one or more networks, and when
922 converted to the target hypervisor you usually want to reconnect those
923 networks at the destination. The options --network, --bridge and --mac
924 allow you to do that.
925 If you are unsure of what networks and bridges are in use on the source
927 hypervisor, then you can examine the source metadata (libvirt XML,
928 vCenter information, etc.). Or you can run virt-v2v with the
929 --print-source option which causes virt-v2v to print out the
930 information it has about the guest on the source and then exit.
931 In the --print-source output you will see a section showing the guest’s
933 Network Interface Cards (NICs):
934 $ virt-v2v [-i ...] --print-source name
936 [...]
937 NICs:
938 Network "default" mac: 52:54:00:d0:cf:0e
939 Bridges are special classes of network devices which are attached to a
941 named external network on the source hypervisor, for example:
942 $ virt-v2v [-i ...] --print-source name
944 [...]
945 NICs:
946 Bridge "br0"
947 To map a specific source bridge to a target network, for example "br0"
949 on the source to "ovirtmgmt" on the target, use:
950 virt-v2v [...] --bridge br0:ovirtmgmt
952 To map every bridge to a target network, use:
954 virt-v2v [...] --bridge ovirtmgmt
956 Fine-grained mapping of guest NICs
958 The --mac option gives you more control over the mapping, letting you
960 map single NICs to either networks or bridges on the target. For
961 example a source guest with two NICs could map them individually to two
962 networks called "mgmt" and "clientdata" like this:
963 $ virt-v2v [...] \
965 --mac 52:54:00:d0:cf:0e:network:mgmt \
966 --mac 52:54:00:d0:cf:0f:network:clientdata
967 Note that virt-v2v does not have the ability to change a guest’s MAC
969 address. The MAC address is part of the guest metadata and must remain
970 the same on source and target hypervisors. Most guests will use the
971 MAC address to set up persistent associations between NICs and internal
972 names (like "eth0"), with firewall settings, or even for other purposes
973 like software licensing.
974 Resource requirements
976 Network
977 The most important resource for virt-v2v appears to be network
979 bandwidth. Virt-v2v should be able to copy guest data at gigabit
980 ethernet speeds or greater.
981 Ensure that the network connections between servers (conversion server,
983 NFS server, vCenter, Xen) are as fast and as low latency as possible.
984 Disk space
986 Virt-v2v places potentially large temporary files in $TMPDIR (which is
988 /var/tmp if you don't set it). Using tmpfs is a bad idea.
989 For each guest disk, an overlay is stored temporarily. This stores the
991 changes made during conversion, and is used as a cache. The overlays
992 are not particularly large - tens or low hundreds of megabytes per disk
993 is typical. In addition to the overlay(s), input and output methods
994 may use disk space, as outlined in the table below.
995 -i ova
997 This temporarily places a full copy of the uncompressed source
998 disks in $TMPDIR.
999 -o glance
1001 This temporarily places a full copy of the output disks in $TMPDIR.
1002 -o local
1004 -o qemu
1005 You must ensure there is sufficient space in the output directory
1006 for the converted guest.
1007 See also "Minimum free space check in the host" below.
1009 VMware vCenter resources
1011 Copying from VMware vCenter is currently quite slow, but we believe
1013 this to be an issue with VMware. Ensuring the VMware ESXi hypervisor
1014 and vCenter are running on fast hardware with plenty of memory should
1015 alleviate this.
1016 Compute power and RAM
1018 Virt-v2v is not especially compute or RAM intensive. If you are
1020 running many parallel conversions, then you may consider allocating one
1021 CPU core and 2 GB of RAM per running instance.
1022 Virt-v2v can be run in a virtual machine.
1024 Trimming
1026 Virt-v2v attempts to optimize the speed of conversion by ignoring guest
1028 filesystem data which is not used. This would include unused
1029 filesystem blocks, blocks containing zeroes, and deleted files.
1030 To do this, virt-v2v issues a non-destructive fstrim(8) operation. As
1032 this happens to an overlay placed over the guest data, it does not
1033 affect the source in any way.
1034 If this fstrim operation fails, you will see a warning, but virt-v2v
1036 will continue anyway. It may run more slowly (in some cases much more
1037 slowly), because it is copying the unused parts of the disk.
1038 Unfortunately support for fstrim is not universal, and it also depends
1040 on specific details of the filesystem, partition alignment, and backing
1041 storage. As an example, NTFS filesystems cannot be fstrimmed if they
1042 occupy a partition which is not aligned to the underlying storage.
1043 That was the default on Windows before Vista. As another example, VFAT
1044 filesystems (used by UEFI guests) cannot be trimmed at all.
1045 fstrim support in the Linux kernel is improving gradually, so over time
1047 some of these restrictions will be lifted and virt-v2v will work
1048 faster.
1049 Post-conversion tasks
1051 Guest network configuration
1052 Virt-v2v cannot currently reconfigure a guest’s network configuration.
1054 If the converted guest is not connected to the same subnet as the
1055 source, its network configuration may have to be updated. See also
1056 virt-customize(1).
1057 Converting a Windows guest
1059 When converting a Windows guests, the conversion process is split into
1061 two stages:
1062 1. Offline conversion.
1064 2. First boot.
1066 The guest will be bootable after the offline conversion stage, but will
1068 not yet have all necessary drivers installed to work correctly. These
1069 will be installed automatically the first time the guest boots.
1070 N.B. Take care not to interrupt the automatic driver installation
1072 process when logging in to the guest for the first time, as this may
1073 prevent the guest from subsequently booting correctly.
1074 Free space for conversion
1076 Free space in the guest
1077 Virt-v2v checks there is sufficient free space in the guest filesystem
1079 to perform the conversion. Currently it checks:
1080 Linux root filesystem or Windows "C:" drive
1082 Minimum free space: 20 MB
1083 Linux /boot
1085 Minimum free space: 50 MB
1086 This is because we need to build a new initramfs for some
1088 Enterprise Linux conversions.
1089 Any other mountable filesystem
1091 Minimum free space: 10 MB
1092 Minimum free space check in the host
1094 You must have sufficient free space in the host directory used to store
1096 temporary overlays (except in --in-place mode). To find out which
1097 directory this is, use:
1098 $ df -h "`guestfish get-cachedir`"
1100 Filesystem Size Used Avail Use% Mounted on
1101 /dev/mapper/root 50G 40G 6.8G 86% /
1102 and look under the "Avail" column. Virt-v2v will refuse to do the
1104 conversion at all unless at least 1GB is available there.
1105 See also "Resource requirements" above.
1107 Running virt-v2v as root or non-root
1109 Nothing in virt-v2v inherently needs root access, and it will run just
1110 fine as a non-root user. However, certain external features may
1111 require either root or a special user:
1112 Mounting the Export Storage Domain
1114 When using -o rhv -os server:/esd virt-v2v has to have sufficient
1115 privileges to NFS mount the Export Storage Domain from "server".
1116 You can avoid needing root here by mounting it yourself before
1118 running virt-v2v, and passing -os /mountpoint instead, but first of
1119 all read the next section ...
1120 Writing to the Export Storage Domain as 36:36
1122 RHV-M cannot read files and directories from the Export Storage
1123 Domain unless they have UID:GID 36:36. You will see VM import
1124 problems if the UID:GID is not correct.
1125 When you run virt-v2v -o rhv as root, virt-v2v attempts to create
1127 files and directories with the correct ownership. If you run
1128 virt-v2v as non-root, it will probably still work, but you will
1129 need to manually change ownership after virt-v2v has finished.
1130 Writing to libvirt
1132 When using -o libvirt, you may need to run virt-v2v as root so that
1133 it can write to the libvirt system instance (ie. "qemu:///system")
1134 and to the default location for disk images (usually
1135 /var/lib/libvirt/images).
1136 You can avoid this by setting up libvirt connection authentication,
1138 see http://libvirt.org/auth.html. Alternatively, use -oc
1139 qemu:///session, which will write to your per-user libvirt
1140 instance.
1141 Writing to Openstack
1143 Because of how Cinder volumes are presented as /dev block devices,
1144 using -o openstack normally requires that virt-v2v is run as root.
1145 Writing to Glance
1147 This does not need root (in fact it probably won’t work), but may
1148 require either a special user and/or for you to source a script
1149 that sets authentication environment variables. Consult the Glance
1150 documentation.
1151 Minimal XML for -i libvirtxml option
1153 When using the -i libvirtxml option, you have to supply some libvirt
1154 XML. Writing this from scratch is hard, so the template below is
1155 helpful.
1156 Note this should only be used for testing and/or where you know what
1158 you're doing! If you have libvirt metadata for the guest, always use
1159 that instead.
1160 <domain type='kvm'>
1162 <name> NAME </name>
1163 <memory>1048576</memory>
1164 <vcpu>2</vcpu>
1165 <os>
1166 <type>hvm</type>
1167 <boot dev='hd'/>
1168 </os>
1169 <features>
1170 <acpi/>
1171 <apic/>
1172 <pae/>
1173 </features>
1174 <devices>
1175 <disk type='file' device='disk'>
1176 <driver name='qemu' type='raw'/>
1177 <source file='/path/to/disk/image'/>
1178 <target dev='hda' bus='ide'/>
1179 </disk>
1180 <interface type='network'>
1181 <mac address='52:54:00:01:02:03'/>
1182 <source network='default'/>
1183 <model type='rtl8139'/>
1184 </interface>
1185 </devices>
1186 </domain>
1187 In-place conversion
1189 It is also possible to use virt-v2v in scenarios where a foreign VM has
1190 already been imported into a KVM-based hypervisor, but still needs
1191 adjustments in the guest to make it run in the new virtual hardware.
1192 In that case it is assumed that a third-party tool has created the
1194 target VM in the supported KVM-based hypervisor based on the source VM
1195 configuration and contents, but using virtual devices more appropriate
1196 for KVM (e.g. virtio storage and network, etc.).
1197 Then, to make the guest OS boot and run in the changed environment, one
1199 can use:
1200 virt-v2v -ic qemu:///system converted_vm --in-place
1202 Virt-v2v will analyze the configuration of "converted_vm" in the
1204 "qemu:///system" libvirt instance, and apply various fixups to the
1205 guest OS configuration to make it match the VM configuration. This may
1206 include installing virtio drivers, configuring the bootloader, the
1207 mountpoints, the network interfaces, and so on.
1208 Should an error occur during the operation, virt-v2v exits with an
1210 error code leaving the VM in an undefined state.
1211 Machine readable output
1213 The --machine-readable option can be used to make the output more
1214 machine friendly, which is useful when calling virt-v2v from other
1215 programs, GUIs etc.
1216 There are two ways to use this option.
1218 Firstly use the option on its own to query the capabilities of the
1220 virt-v2v binary. Typical output looks like this:
1221 $ virt-v2v --machine-readable
1223 virt-v2v
1224 libguestfs-rewrite
1225 colours-option
1226 vdsm-compat-option
1227 input:disk
1228 [...]
1229 output:local
1230 [...]
1231 convert:linux
1232 convert:windows
1233 A list of features is printed, one per line, and the program exits with
1235 status 0.
1236 The "input:" and "output:" features refer to -i and -o (input and
1238 output mode) options supported by this binary. The "convert:" features
1239 refer to guest types that this binary knows how to convert.
1240 Secondly use the option in conjunction with other options to make the
1242 regular program output more machine friendly.
1243 At the moment this means:
1245 1. Progress bar messages can be parsed from stdout by looking for this
1247 regular expression:
1248 ^[0-9]+/[0-9]+$
1250 2. The calling program should treat messages sent to stdout (except
1252 for progress bar messages) as status messages. They can be logged
1253 and/or displayed to the user.
1254 3. The calling program should treat messages sent to stderr as error
1256 messages. In addition, virt-v2v exits with a non-zero status code
1257 if there was a fatal error.
1258 Virt-v2v ≤ 0.9.1 did not support the --machine-readable option at all.
1260 The option was added when virt-v2v was rewritten in 2014.
1261 It is possible to specify a format string for controlling the output;
1263 see "ADVANCED MACHINE READABLE OUTPUT" in guestfs(3).
1264
1266 /usr/share/virtio-win
1267 (Optional)
1268 If this directory is present, then virtio drivers for Windows
1270 guests will be found from this directory and installed in the guest
1271 during conversion.
1272
1274 "TMPDIR"
1275 Location of the temporary directory used for the potentially large
1276 temporary overlay file.
1277 See the "Disk space" section above.
1279 "VIRT_TOOLS_DATA_DIR"
1281 This can point to the directory containing data files used for
1282 Windows conversion.
1283 Normally you do not need to set this. If not set, a compiled-in
1285 default will be used (something like /usr/share/virt-tools).
1286 This directory may contain the following files:
1288 rhsrvany.exe
1290 (Required when doing conversions of Windows guests)
1291 This is the RHSrvAny Windows binary, used to install a
1293 "firstboot" script in the guest during conversion of Windows
1294 guests.
1295 See also: "https://github.com/rwmjones/rhsrvany"
1297 pvvxsvc.exe
1299 This is a Windows binary shipped with SUSE VMDP, used to
1300 install a "firstboot" script in Windows guests. It is required
1301 if you intend to use the --firstboot or --firstboot-command
1302 options with Windows guests.
1303 rhev-apt.exe
1305 (Optional)
1306 The RHV Application Provisioning Tool (RHEV APT). If this file
1308 is present, then RHEV APT will be installed in the Windows
1309 guest during conversion. This tool is a guest agent which
1310 ensures that the virtio drivers remain up to date when the
1311 guest is running on Red Hat Virtualization (RHV).
1312 This file comes from Red Hat Virtualization (RHV), and is not
1314 distributed with virt-v2v.
1315 "VIRTIO_WIN"
1317 This is where virtio drivers for Windows are searched for.
1318 If unset, then we look for drivers in whichever of these paths is
1320 found first:
1321 /usr/share/virtio-win/virtio-win.iso
1323 The ISO containing virtio drivers for Windows.
1324 /usr/share/virtio-win
1326 The exploded tree of virtio drivers for Windows. This is
1327 usually incomplete, hence the ISO is preferred.
1328 ( if unset). It can be a directory or point to virtio-win.iso (CD
1330 ROM image containing drivers).
1331 See "Enabling virtio".
1333 For other environment variables, see "ENVIRONMENT VARIABLES" in
1335 guestfs(3).
1336
1338 virt-v2v-copy-to-local(1)
1339 There are some special cases where virt-v2v cannot directly access
1340 the remote hypervisor. In that case you have to use
1341 virt-v2v-copy-to-local(1) to make a local copy of the guest first,
1342 followed by running "virt-v2v -i libvirtxml" to perform the
1343 conversion.
1344 engine-image-uploader(8)
1346 Variously called "engine-image-uploader", "ovirt-image-uploader" or
1347 "rhevm-image-uploader", this tool allows you to copy a guest from
1348 one oVirt or RHV Export Storage Domain to another. It only permits
1349 importing a guest that was previously exported from another
1350 oVirt/RHV instance.
1351 import-to-ovirt.pl
1353 This script can be used to import guests that already run on KVM to
1354 oVirt or RHV. For more information, see this blog posting by the
1355 author of virt-v2v:
1356 https://rwmj.wordpress.com/2015/09/18/importing-kvm-guests-to-ovirt-or-rhev/#content
1358
1360 virt-p2v(1), virt-customize(1), virt-df(1), virt-filesystems(1),
1361 virt-sparsify(1), virt-sysprep(1), guestfs(3), guestfish(1),
1362 qemu-img(1), virt-v2v-copy-to-local(1), virt-v2v-test-harness(1),
1363 engine-image-uploader(8), import-to-ovirt.pl, nbdkit(1),
1364 nbdkit-vddk-plugin(1), http://libguestfs.org/.
1365
1367 Matthew Booth
1368 Cédric Bosdonnat
1370 Tomáš Golembiovský
1372 Shahar Havivi
1374 Roman Kagan
1376 Mike Latimer
1378 Nir Soffer
1380 Richard W.M. Jones
1382 Pino Toscano
1384 Tingting Zheng
1386
1388 Copyright (C) 2009-2019 Red Hat Inc.
1389
1391 This program is free software; you can redistribute it and/or modify it
1392 under the terms of the GNU General Public License as published by the
1393 Free Software Foundation; either version 2 of the License, or (at your
1394 option) any later version.
1395 This program is distributed in the hope that it will be useful, but
1397 WITHOUT ANY WARRANTY; without even the implied warranty of
1398 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
1399 General Public License for more details.
1400 You should have received a copy of the GNU General Public License along
1402 with this program; if not, write to the Free Software Foundation, Inc.,
1403 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
1404
1406 To get a list of bugs against libguestfs, use this link:
1407 https://bugzilla.redhat.com/buglist.cgi?component=libguestfs&product=Virtualization+Tools
1408 To report a new bug against libguestfs, use this link:
1410 https://bugzilla.redhat.com/enter_bug.cgi?component=libguestfs&product=Virtualization+Tools
1411 When reporting a bug, please supply:
1413 · The version of libguestfs.
1415 · Where you got libguestfs (eg. which Linux distro, compiled from
1417 source, etc)
1418 · Describe the bug accurately and give a way to reproduce it.
1420 · Run libguestfs-test-tool(1) and paste the complete, unedited output
1422 into the bug report.
1423libguestfs-1.40.2 2019-02-07 virt-v2v(1)