When restoring a backup and the storage the disks would be created on
doesn't allow 'images', the process errors without cleanup.
This is the same behaviour we currently have when the storage is
disabled.
Signed-off-by: Matthias Heiserer <m.heiserer@proxmox.com>
Reviewed-by: Fabian Ebner <f.ebner@proxmox.com>
Tested-by: Fabian Ebner <f.ebner@proxmox.com>
preparation for also clamping on hotplug and lower the minimum in the
schema so that the full v2 range can be used.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
when passing a config from one cluster to another, we want to be strict
when parsing - it's better to fail the migration early and upgrade the
target node instead of failing the migration later (when significant
work for transferring disks and/or state has already been done) or not
at all, but silently lose config settings that the target doesn't
understand.
this also might be helpful in other cases - e.g. when restoring from a
backup.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
since we are going to reuse the same mechanism/code for network bridge
mapping and pve-container.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
While existing callers are not using the parameter after the call,
the modification is rather unexpected and could lead to bugs quickly.
Also avoid setting an undef value in the hash, but use delete instead.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
We drop properties which we do not understand and we call
`vmconfig_apply_pending` on stop and before start, so if a user tried
to edit the config or downgraded qemu-server they may get stuff
dropped from the config just by doing a stop/start, which may be a
bit too confusing, also the write is just unnecessary then.
we also have the same skipping logic when starting vms, this way we
avoid calling 'write_config' when there are no present changes to
commit.
Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
The volid may change if local-storage migration is involved, we need
to tell the target node the new one and update the in-memory config
for starting the target VM accordingly.
Reported here: https://forum.proxmox.com/threads/99906/#post-431345
this possibly breaks migration new -> old iff
- spice is not used (else the explicit ticket wins because it comes
later)
- a local TPM state volume is used
- that local TPM state volume has a different volume id on the target
node (switched storage, volname already taken, ..)
because the target node will then mis-interpret the tpmstate0 line as
spice ticket and set it accordingly. if the old tpm state volume ID does
not exist on the target node, migration will fail. if it exists by
chance, it might work albeit with a wrong spice ticket (new because of
this patch) and tpm state volume (pre-existing breakage).
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
This patch fixes the wrong attempt of setting up an NBD server for
the replicated TPM state volume, in contrast to the other volumes the
TPM state is managed by swtpm and isn't available to QEMU for
block-migration/bitmap tracking.
Note that we do migrate the state volume via a storage migration
anyway if necessary.
This code path was only triggered for replicated VMs with TPM.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
else a user cannot use more than one mdev per card per host.
We do not need to reserve them at all, since sysfs will error out
on creation/reuse anyway
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
issue reported in community forum [0][1], like "serial[n]" display we
also need to set this option for "none", otherwise we get a boot
loop.
[0]: https://forum.proxmox.com/threads/99508
[1]: https://forum.proxmox.com/threads/97310/post-427129
Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
Reviewed-by: Fabian Ebner <f.ebner@proxmox.com>
Tested-by: Fabian Ebner <f.ebner@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
See commit 17858a1695 (hw/acpi/ich9: Set ACPI PCI hot-plug as default
on Q35)[0] in upstream QEMU repository for details about why the change
was made.
As that change affects systemds predictable interface naming[1],
e.g., by going from a previously `ens18` name to `enp6s18`, it may
have rather bad effects for users that did not setup some .link files
to enforce a specific naming by an more stable information like the
NIC's MAC-Address
The alternative would be making the preferred mode of hotplug an
option like `hotplug-mode=<acpi|pcie>`, but it does not seems like
one would like to change that much in the first place...
Note the changes to the tests and especially the tests with q35
machines that did not change.
[0]: https://gitlab.com/qemu-project/qemu/-/commit/17858a1695
[1]: https://www.freedesktop.org/software/systemd/man/systemd.net-naming-scheme.html#Naming
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Acked-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Reviewed-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Dominik Csapak <d.csapak@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
This is intended to be used to apply some workarounds for the
non-windows ostyped VMs which we'd still like to not pin on a
specific machine version, as normally Linux et al. can cope with such
changes on fresh boot just fine and until now this was a once every
few year issue (albeit systemd's "predictable" interface naming has
some potential to pick up on churn frequency).
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Acked-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Reviewed-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Dominik Csapak <d.csapak@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
currently we only add the creation time (ctime), that was requested
as low priority wish from some users from time to time.
Note that the meta info is not available in the update API endpoints,
and at the moment the code should not change/add/delete it either in
any place.
We may want to update in on actions like clone or backup-restore in
the future, e.g., to also save the time of that event and possibly
the original source VMID, put that can be thought out later.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Acked-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Reviewed-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Dominik Csapak <d.csapak@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
while perl returns the (scalar) result of the last expression
automatically its still nicer to explicitly do so..
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
this allows a user to set a drive to 'read-only'. This can be useful
if a disk should not be written to, or if the backing file/source is
not writable (like a mapped pbs backup to /dev/loopX).
the option is named 'ro', to achieve consistency with containers
while this could also be achieved by setting 'snapshot=1', this would
create a temporary file in /var/tmp which can get quite big.
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
ovmf with SMM enabled will not boot on i440fx (hangs on graphics
initialization), so load the non SMM variant.
should be no issue regarding live-migration since it never worked with
this anyway.
adapts the test and adds one with q35
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
Reviewed-by: Stefan Reiter <s.reiter@proxmox.com>
fix the classic indentation error on `additionalProperties` in the
main QEMU API
drop some not so useful empty lines to avoid making rather huge
methods even bigger (more intimidating, less on screen to grasp the
full picture).
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
swtpm may take a little bit to daemonize, so the pidfile might not be
available right after run_command. Causes an ugly warning about using an
undefined value in a match, so wait up to 5s for it to appear.
Note that in testing this loop only ever got to the first or second
iteration, so I believe the timeout duration should be more than enough.
Also add a missing 'usleep' import, 'usleep' was used before but never
imported, apparently the other case never got triggered...
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
QEMU doesn't know about the tpmstate, so 'do_snapshots_with_qemu' should
never return true in that case. Note that inconsistencies related to
snapshot timing do not matter much, as the actual TPM data is exported
together with other device state by QEMU anyway.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
`properties` is a bit ambiguous and as we have scope and start
runtime properties in the same scope it's good to avoid that
ambiguity.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
on vm start, we reserve all pciids that we use, and
remove the reservation again in vm_stop_cleanup
first with only a time-based reservation but after the vm is started,
we reserve again but with the pid.
for this, we have to move the start_timeout calculation above the
hostpci handling.
also moved the pci initialization out of the conf parsing loop
so that we can reserve all ids before we actually touch any of them
while touching the lines, fix the indentation
this way, when a vm starts with a pci device that is already configured
for a different running vm, will not be started and the user gets
the error that the device is already in use
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Provide support for secure boot by using the new "4m" and "4m-ms"
variants of the OVMF code/vars templates. This is specified on the
efidisk via the 'efitype' and 'ms-keys' parameters.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Starts an instance of swtpm per VM in it's systemd scope, it will
terminate by itself if the VM exits, or be terminated manually if
startup fails.
Before first use, a TPM state is created via swtpm_setup. State is
stored in a 'tpmstate0' volume, treated much the same way as an efidisk.
It is migrated 'offline', the important part here is the creation of the
target volume, the actual data transfer happens via the QEMU device
state migration process.
Move-disk can only work offline, as the disk is not registered with
QEMU, so 'drive-mirror' wouldn't work. swtpm itself has no method of
moving a backing storage at runtime.
For backups, a bit of a workaround is necessary (this may later be
replaced by NBD support in swtpm): During the backup, we attach the
backing file of the TPM as a read-only drive to QEMU, so our backup
code can detect it as a block device and back it up as such, while
ensuring consistency with the rest of disk state ("snapshot" semantic).
The name for the ephemeral drive is specifically chosen as
'drive-tpmstate0-backup', diverging from our usual naming scheme with
the '-backup' suffix, to avoid it ever being treated as a regular drive
from the rest of the stack in case it gets left over after a backup for
some reason (shouldn't happen).
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
if a volume is only referenced in the pending section of a config it was
previously not removed when removing the VM, unless the non-default
'remove unreferenced disks' option was enabled.
keeping track of volume IDs which we attempt to remove gets rid of false
warnings in case a volume is referenced both in the config and the
pending section, or multiple times in the config for other reasons.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
the assumption that the index of the controller matches that of the last
removed drive only holds for virtio-scsi-single controller, which makes
the old code print a warning when removing the last drive of a
non-virtio-scsi-single controller except when the indices line up by
chance.
we can simply only call a simplified qemu_iothread_del when removing a
scsi disk of a VM with the virtio-scsi-single controller, and skip the
call for the other controllers which don't support io-threads anyway.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
The content of the ISO should be the same on both nodes, so offline
migrate the ISO, but don't regenerate it on VM start on the target node.
This way even with snippets the content will not change during live
migration.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
there may be a kernel issue or a bug in how QEMU uses io_uring, but
we have users that report crashes which f.ebner could see on some
workloads, not really deterministic though and it seems that in newer
kernel versions (5.12+) the crash becomes a hang
While we're closing in on the actual issue here (which could be the
same as for RBD) let's disable io_uring for LVM.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
That bit of code seems to be enough here, tested with
qm set VMID --net1 e1000e=EA:93:42:22:10:D8,bridge=vmbr0
on a Alpine Linux and a Windows Server 2016 VM.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
In v2 the range is [1, 10000], but the API allows the old limits from
2 to 262144, so clamp the upper for v2.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
The efidisk never got restored correctly before, since we don't use the
generic print_drive_commandline_full for it, and as such it didn't get a
backing image attached. This not only causes the efidisk data to be lost
on restore, but also an error at the end, since we try to remove a
non-existing PBS blockdev.
Since it is attached differently to a regular drive, adding PBS backing
would be more difficult, but not to worry: an efidisk is small enough
that it doesn't hurt performance to just restore it via the regular
mechanism before starting the VM, and simply excluding it from the live
restore entirely.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
otherwise it'll produce a whole lot of checksum errors
and while this would be nice as a storage feature check,
it's hard to be 100% accurate there anyway since a directory
storage can point anywhere, like for instance a btrfs
directory, causing the same issue...
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
this allows effectively setting ALL volumes as read-only, even if the
disk controller does not support it. without it, IDE and SATA disks
with (base) volumes which are marked read-only/immutable on the storage
level prevent the template VM from starting for backup purposes.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
otherwise backups of templates using UEFI fail with storages like LVM
thin, where the volumes are not writable. disk controllers like IDE and
SATA that don't support being read-only are still broken for UEFI.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
[ drop the readonly=off when not required, resolve merger conflict
from Dominik's EFI disk cache mode fix ]
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
KillMode 'none' is deprecated, and systemd loudly complains about that
in the journal. To avoid the warning, but keep the behaviour the same,
use KillMode 'process'.
This mode does two things differently, which we have to stop it from
doing:
* it sends SIGTERM right when the scope is cancelled (e.g. on shutdown)
-> but only to the "root" process, which in our case is the worker
instance forking QEMU, so it is already dead by the time this happens
* it sends SIGKILL to *all* children after a timeout
-> can be avoided by setting either SendSIGKILL to false, or
TimeoutStopUSec to infinity - for safety, we do both
In my testing, this replicated the previous behaviour exactly, but
without using the deprecated 'none' mode.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
The 'aio' setting is not visible to the guest, and so can be changed
during migrations or snapshots without issue. It is thus only
dependendent on the actual QEMU version being >= 6.0, not machine
version.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
and use it for the vdisk_list call too. This avoids scanning (and picking up
volumes from!) storages that are not even configured to hold images.
Previously, the content type was only enforced when a storage map was present.
Also serves a bit as a preparation to enforce content type on guest startup,
because now migration failure happens early and not only when trying to start
the guest on the remote node.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
storage_check_enabled simply checks for the 'disable' option and then calls
storage_check_node.
While not strictly necessary for a second call where only the storage differs,
e.g. in case of clone, it is more future-proof: if support for a target storage
is added at some point, it might be easy to miss adapting the call.
For the migration checks, the situation is improved by now always catching
disabled (target) storages.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
to avoid potential problems with stringified numbers in Javascript and
elsewehere.
The vmid was not always an integer as the return schema expects, namely
when there was an opt_vmid argument, because the 'ne' comparision coerced the
vmid to be a string then.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Reported in the community forum[0].
In QEMU's hw/scsi/vmw_pvscsi.c in the SCSIBusInfo struct, the max_lun property
is set to 0. This means that in our stack, one cannot have multiple disks and
use 'scsihw: pvscsi' currently, as kvm would fail with
bad scsi device lun: 1
Instead of increasing the lun number, increase the scsi-id, as we already do for
lsi.* (in hw/scsi/lsi53c895a.c the max_lun property is also 0).
[0]: https://forum.proxmox.com/threads/kvm-bad-scsi-device-lun-1.84318/
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Reviewed-by: Stefan Reiter <s.reiter@proxmox.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
on slower ceph clusters, the write pattern of the ovmf booting process
slows down the boot of the vm, so we turn on caching by default
it seems no other storage (until now) behaves like this. if it does in
the future, we can still add them too, or add a 'cache' property for
the efidisk
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
The only caller that didn't use 'images' was removed as part of the migration
refactoring in commit 62a4c963b8, so this is not
even a breaking change as the 'PVE 7' comment might've suggested.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Reviewed-by: Stefan Reiter <s.reiter@proxmox.com>
To bring it better in line with regular restore, also log the
repository, the snapshot and the target for each drive.
While at it, adjust capitalization of existing log line and clean up
repeated '$1' use.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
It's arguably not likely in practice that only an unused volume is still in use
as a base image, but do it for completeness sake.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
QEMU warns us about this:
kvm: -chardev socket,id=qmp,path=/var/run/qemu-server/100.qmp,server,nowait: warning: short-form boolean option 'server' deprecated
Please use server=on instead
kvm: -chardev socket,id=qmp,path=/var/run/qemu-server/100.qmp,server,nowait: warning: short-form boolean option 'nowait' deprecated
Please use wait=off instead
kvm: -vnc unix:/var/run/qemu-server/100.vnc,password: warning: short-form boolean option 'password' deprecated
Please use password=on instead
The new syntax is backwards compatible to at least QEMU 4.0.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
for IDE and SATA, setting the whole drive into readonly mode is not
possible. skip the readonly flag for such drives as a workaround until
we find a better solution.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
except for migration, where it would be subtly backwards-incompatible. Since
there is a scan_volids call for migration, we can't default to filtering in
scan_volids just yet.
Also allows to get rid of the existing filtering hack in rescan().
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Pinned machine versions like "pc-i440fx-4.2+pve2.pxe" would otherwise
get a second "+pve0" suffix, which is incorrect.
Also deal with non-pve pinned versions correctly, i.e.
"pc-i440fx-5.2.pxe" becomes "pc-i440fx-5.2+pve0.pxe".
Handle .pxe suffixes in Machine.pm as well, and add two test cases.
Co-developed-by: Luca Berneking <luca@berneking.net>
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
If, why ever, got "not-ready" again we'd log again the next round.
Improves the behavior for multiple disks, especially on migration
where we mirrored the local disks one by one, but kept reporting on
prev. ones.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
orient on the backup output which got reworked for PVE 6.2/6.3
Avoid overwhelming the user with redundant information, and use human
readable units.
before:
> restore-drive-scsi5: transferred: 167772160 bytes remaining: 8422162432 bytes total: 8589934592 bytes progression: 1.95 % busy: 1 ready: 0
after:
> restore-drive-scsi0: transferred 720.0 MiB of 32.0 GiB (2.20%) in 12s
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Similar to backups, prevent QEMU from being killed by qmeventd during
the live-restore, so a guest can shut itself down without aborting the
restore operation.
Note that the 'close' is only to be explicit, the handle will also be
closed in case an operation errors (i.e. when the 'eval' is left).
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Enables live-restore functionality using the 'alloc-track' QEMU driver.
This allows starting a VM immediately when restoring from a PBS
snapshot. The snapshot is mounted into the VM, so it can boot from that,
while guest reads and a 'block-stream' job handle the restore in the
background.
If an error occurs, the VM is deleted and all data written during the
restore is lost.
The VM remains locked during the restore, which automatically prohibits
any modifications to the config while restoring. Some modifications
might potentially be safe, however, this is experimental enough that I
believe this would cause more bad stuff(tm) than actually satisfy any
use cases.
Pool handling is slightly adjusted so the VM can be added to the pool
before the restore starts.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Uses the custom 'alloc-track' filter node to redirect writes to the
original drives target, while unwritten blocks will be read from the
specified PBS snapshot.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
...so it works with other block jobs as well. Intended use case is
block-stream, which also requires a new "auto" (wait only) completion
mode, since it finishes automatically anyway.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
cloud-init's SLAAC option was disabled in 2018 because there was no
support for it. Now that cloud-init 19.4 or newer versions are more
widespread, we can finally reenable it.
Also include minimum required cloud-init version for SLAAC support in
format description.
Tested on Ubuntu 20.04 (ci 20.4), CentOS 8 (ci 19.4), Debian 10 (ci
20.2).
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
In testing this usually completes almost immediately, but in theory this
is a storage/IO operation and as such can take a bit to finish. It's
certainly not unthinkable that it might take longer than the default *3
seconds* we've given it so far. Make it a minute.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
A "savevm" call (both our async variant and the upstream sync one) use
migration code internally. As such, they both expect migration
capabilities to be set.
This is usually not a problem, as the default set of capabilities is ok,
however, it leads to differing snapshot settings if one does a snapshot
after a machine has been live-migrated (as the capabilities will persist
from that), which could potentially lead to discrepencies in snapshots
(currently it seems to be fine, but it still makes sense to set them to
safeguard against future changes).
Note that we do set the "dirty-bitmaps" capability now (if
query-proxmox-support reports true), which has three effects:
1) PBS dirty-bitmaps are preserved in snapshots, enabling
fast-incremental backups to work after rollback (as long as no newer
backups exist), including for hibernate/resume
2) snapshots taken from now on, with a QEMU version supporting bitmap
migration, *might* lead to incompatibility of these snapshots with
QEMU versions that don't know about bitmaps at all (i.e. < 5.0 IIRC?)
- forward compatibility is still given, and all other capabilities we
set go back to very old versions
3) since we now explicitly disable bitmap saving if the version doesn't
report support, we avoid crashes even with not-updated QEMU versions
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
At this stage, there are no keys in %storage_limits to iterate over. The
refactoring in commit 9f3d73bc35 broke the logic
by accident.
Also explicitly set zero if there is no limit to avoid repeating the
get_bandwith_limit call for the same storage. When accessing the value later,
zero is already correctly handled as 'no limit'.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
and use file_set_contents to really commit it afterwards. Mostly done as a
preparation for the later patch for sanitizing the config on restore, but
shouldn't hurt by itself either.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
A fix for violating a important standard for booting[0] in recently
packaged QEMU 5.2 surfaced some issues with Windows based VMs in our
forum[1], which seem to be quite sensitive for such changes (it seems
they derive lots of their device assignment from ACPI).
User visible effects are loss of any network configuration due to
windows thinking it was swapped with a new one, and starts with a
fresh config - this is mostly problematic for setups with static
address assignment.
There may be lots of other, more subtle, effects and the PVE admin is
also not always the VM admin, so we really need to avoid such
negative effects. Do this by pinning the version of any windows based
VMs to either the minimum of (5.1, kvm-version) for existing VMs or
the kvm-version at time of VM creation for new ones.
There are patches in pve-manager for user to be able to change the
pinned version themself in the webinterface, so this can now also get
adapted more easily if there surface any other issues (with new or
old version) in the future.
0: https://lists.gnu.org/archive/html/qemu-devel/2021-02/msg08484.html
1: https://forum.proxmox.com/threads/warning-latest-patch-just-broke-all-my-windows-vms-6-3-4-patch-inside.84915/page-2#post-373331
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Moving to Ceph is very slow when bs=1. Instead, use a larger block size in
combination with the (currently) PVE-specific osize option to specify the
desired output size.
Suggested-by: Dietmar Maurer <dietmar@proxmox.com>
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
currently only pending changes are applied when we regenerate
image on a running vm, but not the pending delete.
Signed-off-by: Alexandre Derumier <aderumier@odiso.com>
by checking if the vm is paused at the beginning and skipping the
resume now we also skip the qga freeze/thaw (which cannot work if the
vm is paused)
moved the 'vm_is_paused' sub from the api to PVE/QemuServer.pm so it
is available everywhere we need it.
since a suspend backup would pause the vm anyway, we can skip that
step also
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Fabian Ebner <f.ebner@proxmox.com>
this was previously covered by the "lets destroy ever disk which
matches the VMID" feature we disarmed a bit.
As unused disks are referenced in the config, it is not subtle to
destroy them (and we always did in the past) so fix that regression
again for explicitly referenced but unused disks.
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Since an old change released with a version bump on 2009-09-07, we
search all enabled storages for VMID maching volumes on VM removal
and purge those too.
This has multiple pitfalls and may be quite unexpected for some
users.
It can make problems when:
* on recovery a VM is created, before disks are reattached the admin
notices some settings issues and chooses to just recreate the VM;
but during destroying the dummy VM all related disks get destroyed
unconditionally which may result in data loss. This actually
happened and is the original reason for the decision to change
this.
* a storage is shared between PVE instance (between a set of clusters
and/or single nodes), while this is against our rules it may still
come as a surprise if destroying a VM on node A may destroy
unrelated and unreferenced disks on the unrelated node B without
asking or allowing to avoid that.
As this the removal of matching but unreferenced disks can result in
permanent data loss (up to the last backup) and may be to subtle and
unforgiving, allow to opt-out of it.
In the long run we want to make this opt-in, but that is an API
change and so needs to wait for next major release. But, we can adapt
the GUI already to make it opt-in there, catching most of the cases.
side-note: CT do not have this behavior at all
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
On clone_vm when cloning the disks while the VM is running, we use
drive-mirror. We skip completion until the last disk, but with a
cloudinit disk there's no drive-mirror and so no completion done. If it
is the last disk in the hash, we never complete the drive-mirror jobs
and no further cloning is possible as there are already active jobs
using the disks.
To fix it we have to call qemu_drive_mirror_monitor directly in the case
of cloudinit when completion is requested and there are jobs defined.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
The phrasing left some room for speculation when this would be triggered.
E.g. after cloning a full VM?
Currently the only instances where it is used is when a disk is moved or
a VM migrated.
Signed-off-by: Aaron Lauterer <a.lauterer@proxmox.com>
We only added the format extension when it was not 'raw'. But on file level
storages we always require it. To fix this, always add the format
extension if the storage provides the 'path' property.
This is the same logic we use in create_disks for cloudinit disks.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
by partially reverting 4df98f2f14 and fixing the
line-length issue differently. The commit didn't update two later usages of
$size, breaking copying the efidisk. The other usage as a parameter to
qemu_img_convert() is luckily only cosmetic, for progress output.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
this fixes the issue that we did not generate the correct repository
url for pbs storages that contained an ipv6 address or a port
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Extends print_recursive_hash for the CLI to handle JSON booleans so the
result will actually show up in 'qm status --verbose'.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
When the VM is in status 'shutdown', i.e. after the guest issues a
powerdown while a backup is running, QEMU requires a 'system_reset' to
be issued before 'cont' can boot the guest again.
Additionally, when the VM has been powered down during a backup, the
logically correct call would be a 'vm_start', so automatically vm_resume
from vm_start in case this situation occurs. This also means the GUI can
cope with this almost unchanged.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Ignore shutdowns triggered from within the guest in favor of detecting
them via qmeventd and stopping the QEMU process that way.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
by adding the missing argument (otherwise all the other ones are shifted
one slot to the left, which is of course bogus).
this has been broken since 2018 (d559309), but was only made
visible/caused a failure with the recent changes adding
use strict;
use warnings;
to PVE::QemuServer::PCI
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
We query QEMU if it's safe before enabling it, as on versions without
the necessary patches it not only would be useless, but can actually
lead to hangs.
PBS state is always migrated, as it's a small amount of data anyway, so
we don't need to set a specific flag for it.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Specifying 'boot: order=' was intended to be used for an empty bootorder
(i.e. no boot devices), but as it turns out our format parser doesn't
like empty '-list' properties if they are nested in a subformat.
Fixing this in JSONSchema sounds like a risky move, so instead just
write 'boot: ' (without 'order=') to indicate an empty bootorder. The
rest of the code handles it just fine, as this was valid before too.
Incidentally also fixes a bug where you couldn't create a new VM without
any disks if no explicit 'boot' property was specified (i.e. a simple
'qm create 100' without any parameters would fail).
Reported-by: Dominic Jäger <d.jaeger@proxmox.com>
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
fixes commit 74c17b7a23 which moved
this code here, but forgot to pass $vga ref, as the module was not
using warning nor strict mode this was not caught..
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
After migration or a rollback the cloudinit disk might not be allocated, so
volume_size_info() fails. As we override the value anyway for cloudinit
and efi disks simply move the volume_size_info() call into the 'else'
branch.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
(also fixes#3011)
Deprecates the old-style 'boot' and 'bootdisk' options by adding a new
'order=' subproperty to 'boot'.
This allows a user to specify more than one disk in the boot order,
helping with newer versions of SeaBIOS/OVMF where disks without a
bootindex won't be initialized at all (breaks soft-raid and some LVM
setups).
This also allows specifying a bootindex for USB and hostpci devices,
which was not possible before. Floppy boot support is not supported in
the new model, but I doubt that will be a problem (AFAICT we can't even
attach floppy disks to a VM?).
Default behaviour is intended to stay the same, i.e. while new VMs will
receive the new 'order' property, it will be set so the VM starts the
same as before (using get_default_bootorder).
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
The format is unused in this commit, but will replace the current
string-based format of the 'boot' property. It is included since the
parameter of bootorder_from_legacy follows it.
Two helper methods are introduced:
* bootorder_from_legacy: Parses the legacy format into a hash closer to
what the new format represents
* get_default_bootdevices: Encapsulates the legacy default behaviour if
nothing is specified in the boot order
resolve_first_disk is simplified and gets a new $cdrom parameter to
control the behaviour of excluding CD-ROMs or instead searching for only
them.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
We already keep hugepages if they are created with the kernel
commandline (hugepagesz=x hugepages=y), but some setups (specifically
hugepages across multiple NUMA nodes) cannot be configured that way.
Since we always clear these hugepages at VM shutdown, rebooting a VM
that uses them might not work, since the requested count might not be
available anymore by the time we want to use them (also, we would then
no longer allocate them correctly on the NUMA nodes).
Add a 'keephugepages' parameter to skip cleanup and simply leave them
untouched.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
until we maybe have a 'pbs-backup' that links Qemu and PBS like
'pbs-restore', we need to do a regular backup for the template case to
support all storage types and image formats.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Otherwise a warning is printed if the bios is not set in the config.
reported via community forum:
https://forum.proxmox.com/threads/warning-in-qemuserver.74683/
reproduced and tested that the patch fixes the issue.
Signed-off-by: Stoiko Ivanov <s.ivanov@proxmox.com>
Edid support was added with Qemu 5. Windows guests seem to not be able
to get all possible resolutions if the default std VGA device is used as
GPU and the VM boots in BIOS mode. The result is that only one of the
following three resolutions can be configured:
800x600
1024x768
1920x1080
It is important to note that just booting a Windows VM with the edid=off
parameter will not make the large list of resolutions available. It
seems that Windows is caching the list of possible resolutions
somewhere [0].
Uninstalling the 'Microsoft Basic Display Adapter' in the device manager
and rebooting the VM is one way I found to force Windows to recreate the
list of possible resolutions.
Signed-off-by: Aaron Lauterer <a.lauterer@proxmox.com>
[0] https://lists.nongnu.org/archive/html/qemu-devel/2020-07/msg07128.html
In config_to_command, '-loadstate' will be added whenever there is a
vmstate in the config. But in vm_start_nolock, the resume parameter
is used to calculate the appropriate timeout and to remove the vmstate
after the start. The resume parameter was only set if there is a
'suspended' lock, but apparently [0] we cannot rely on the lock to be
set if and only if there is a vmstate.
[0]: https://forum.proxmox.com/threads/task-error-start-failed.72450
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
during refactoring, the vmid got lost, but is necessary to get
the correct mdev id
Fixes commit 74c17b7a23
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
[ reference fixed commit ]
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
pbs-restore might not stay there like that forever and if
this code path changes we should remember to also remove the
environment variables
Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
Pass new size directly, so the function doesn't need to know about
how some hash is organized. And return a message directly, instead
of both size-strings. Also dropped the wantarray, because both
existing callers use the message anyways.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Legacy IGD passthrough requires address 00:1f.0 to not be assigned to
anything on QEMU startup (currently it's assigned to bridge pci.2).
Changing this in general would break live-migration, so introduce a new
hostpci parameter "legacy-igd", which if set to 1 will move that bridge
to be nested under bridge 1.
This is safe because:
* Bridge 1 is unconditionally created on i440fx, so nesting is ok
* Defaults are not changed, i.e. PCI layout only changes when the new
parameter is specified manually
* hostpci forbids migration anyway
Additionally, the PT device has to be assigned address 00:02.0 in the
guest as well, which is usually used for VGA assignment. Luckily, IGD PT
requires vga=none, so that is not an issue either.
See https://git.qemu.org/?p=qemu.git;a=blob;f=docs/igd-assign.txt
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
netdev_add is now a proper qmp command, which means that it verifies
the parameter types properly
instead of sending strings, we now have to choose the correct
types for the parameters
bool for vhost
and uint64 for queues
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
the special case was dropped when moving this to pve-storage.
fixes commit c6d517835a
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
fixes commit 940e2a3a06
QEMU 4.1 will fail to start a guest with an audio device set with:
> Property '.audiodev' not found
Signed-off-by: Aaron Lauterer <a.lauterer@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
If /dev/hwrng exists, but no actual generator is connected (or it is
disabled on the host), QEMU will happily start the VM but crash as soon
as the guest accesses the VirtIO RNG device.
To prevent this unfortunate behaviour, check if a useable hwrng is
connected to the host before allowing the VM to be started.
While at it, clean up config_to_command by moving new and existing rng
source checks to a seperate sub.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
It's possible to have a VM with OVMF but without an efidisk, so don't
call parse_drive on a potential undef value.
Partial revert of 818c3b8d91 ("cfg2cmd: ovmf: code cleanup")
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
lock_file is used by PVE::QemuServer::Memory, but it does properly 'use
PVE::Tools ...' itself so we can drop them in the main module.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
regex to reduce the code duplication, as archive_info and
decompressor_info provides the same information as well.
Signed-off-by: Alwin Antreich <a.antreich@proxmox.com>
and refactor the test_volid closure. Like this get_replicatable_volumes doesn't
need a separate loop for unused volumes anymore. For get_vm_volumes, which is used
for activation/deactivation of volumes at migration and deactivation in vm_stop_cleanup,
includes those volumes now. For migration it's an improvement, because those volumes
might need to be migrated and for vm_stop_cleanup it shouldn't hurt. The last user
of foreach_volid is check_vm_disks_local used by migrate_vm_precondition,
where information about the additional volumes doesn't hurt either.
Note that replicate is (still) set by default, so the behavior for
get_replicatable_volumes for unused volumes should not change.
Hibernation vmstate files are now also included and recognized as 'is_vmstate'.
The 'size' attribute will not be overwritten by subsequent iterations for the
same volid anymore (a volid may appear both in the config and in snapshots),
so the size from the current config is now preferred.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
when a backup includes a cloudinit disk on a non-existent storage,
the restore fails with 'storage' does not exist
this happens because we want to get the format of the disk, by
checking the source storage
we fix this by using the target storage first and only the source as
fallback
this will still fail if neither storage exists
(which is ok, since we cannot restore then anyway)
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
With Qemu 4.2 a new `audiodev` property was introduced [0] to explicitly
specify the backend to be used for the audio device. This is accompanied
with a warning that the fallback to the default audio backend is
deprecated.
[0] https://wiki.qemu.org/ChangeLog/4.2#Audio
Signed-off-by: Aaron Lauterer <a.lauterer@proxmox.com>
This makes it possible to migrate a VM with volumes store1:vm-123-disk-0
store2:vm-123-disk-0 to some targetstorage. Also prevents migration failure
when there is an orphaned disk with the same volid on the target.
To avoid confusion, the name should not change for 'vmstate'-volumes.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
It was necessary to move foreach_volid back to QemuServer.pm
In VZDump/QemuServer.pm and QemuMigrate.pm the dependency on
QemuConfig.pm was already there, just the explicit "use" was missing.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Upstream marks these as having a micro-version of >=90, unfortunately the
machine versions are bumped earlier so testing them is made unnecessarily
difficult, since the version checking code would abort on migrations etc...
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
[ Thomas: do so refactor ]
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
so that pve-container and qemu-server use the same one, in preparation
for moving it to JSONSchema and having a bridgepair format.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Just like with live-migration, custom CPU models might change after a
snapshot has been taken (or a VM suspended), which would lead to a
different QEMU invocation on rollback/resume.
Save the "-cpu" argument as a new "runningcpu" option into the VM conf
akin to "runningmachine" and use as override during rollback/resume.
No functional change with non-custom CPU types intended.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
This is required to support custom CPU models, since the
"cpu-models.conf" file is not versioned, and can be changed while a VM
using a custom model is running. Changing the file in such a state can
lead to a different "-cpu" argument on the receiving side.
This patch fixes this by passing the entire "-cpu" option (extracted
from /proc/.../cmdline) as a "qm start" parameter. Note that this is
only done if the VM to migrate is using a custom model (which we can
check just fine, since the <vmid>.conf *is* versioned with pending
changes), thus not breaking any live-migration directionality.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
in addition to printing it. preparation for remote cluster migration,
where we want to return this in a structured fashion over the migration
tunnel instead of parsing stdout via SSH.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
into one sub that retrieves the local disks, and the actual NBD
allocation. that way, remote incoming migration can just call the NBD
allocation with a custom list of volume names/storages/..
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
the syntax is backwards compatible, providing a single storage ID or '1'
works like before. the new helper ensures consistent behaviour at all
call sites.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
as preparation of targetstorage mapping and remote migration. this also
removes re-using of the $local_volumes hash in the original code.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
to start breaking up vm_start before extending parts for new migration
features like storage and network mapping.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
as preparation for refactoring it further. remote migration will add
another 1-2 parameters, and it is already unwieldly enough as it is.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
on storages where the minimum size of images is bigger than the real
OVMF_VARS.fd file, they get padded to their minimum size
when using such an image, qemu maps it fully to the vm, but the efi
does not find the vars region and creates a file on the first efi
partition it finds
this breaks some settings in the ovmf, such as resolution
to fix this, we have to specify the size for the pflash, so that
qemu only maps the first n bytes in the vm (this only works for
raw files, not for qcow2)
we also have to use the correct size when converting between storages
in 'clone_disk' (used for move disk and cloning vms) and when
live migrating to different storages
when we now expect that the source image is always correctly used/created
(e.g. raw with size=x in pflash argument) then we always create the
target correctly
when encountering users which have a non-valid image (e.g. a efidisk
moved from zfs to qcow2 before this patch), we have to tell them to
recreate the efidisk and the settings on it
we have to version_guard it to 4.1+pve2 (since we haven't bumped yet
since the change to pve2)
also add 2 tests, one for the old version and one for the new
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
Reviewed-by: Stefan Reiter <s.reiter@proxmox.com>
[ Thomas: rebased to master ]
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
by only checking for replicatable volumes when a replication job is
defined, and passing only actually replicated volumes to the target node
via STDIN, and back via STDOUT.
otherwise this can pick up theoretically replicatable, but not actually
replicated volumes and treat them wrong.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
$cpu_fmt is being reused for custom CPUs as well as VM-specific CPU
settings. The "pve-vm-cpu-conf" format is introduced to verify a config
specifically for use as VM-specific settings.
"pve-cpu-conf" is registered for use in custom CPU API calls (where no
additional checks are required).
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
There is a need to set $noerr, because otherwise migration for a
VM with a non-replicatable volume fails with:
missing replicate feature on volume 'myfs:107/vm-107-disk-2.raw'
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
with incremental drive-mirror and dirty-bitmap tracking.
1.) get replicated disks that are currently referenced by running VM
2.) add a block-dirty-bitmap to each of them
3.) replicate ALL replicated disks
4.) pass bitmaps from 2) to drive-mirror for disks from 1)
5.) skip replicated disks when cleaning up volumes on either source or
target
added error handling is just removing the bitmaps if an error occurs at
any point after 2, except when the handover to the target node has
already happened, since the bitmaps are cleaned up together with the
source VM in that case.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
by re-using a dirty bitmap that represents changes since the divergence
of source and target volume. requires a qemu that supports incremental
drive-mirroring, and will die otherwise.
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Tested-by: Stefan Reiter <s.reiter@proxmox.com>
Moved code so that initialization of drivedesc_hash stays a single block.
Avoid auto-vivication in parse_drive.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
E.g.: If a feature requires 4.1+pveN and we're using machine version 4.2
we don't need to increase the pve version to N (4.2+pve0 is enough).
We check this by doing a min_version call against a non-existant higher
pve-version for the major/minor tuple we want to test for, which can
only work if the major/minor alone is high enough.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
As the NBD server spawned by qemu can only listen on a single socket,
we're dependent on a version being passed to vm_start that indicates
which protocol can be used, TCP or Unix, by the source node.
The change in socket type (TCP to Unix) comes with a different URI. For
unix sockets it has the form: 'nbd:unix:<path/to/socket>:exportname=<device>'.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
With Qemu 4.2 we encountered a problem with unix sockets and SSH socket
forwarding for drive-mirror. It seems the socket gets reopened again and
again after it closes for some reason. This can be worked around by
specifying 'block-job-cancel' instead of 'block-job-complete' when we're
not interested in swapping the disks again from NBD to their original
protocol. This is always the case when we use drive-mirror for live
migrating a VM.
qemu_drive_mirror is used for migration and for clone_disk. All in all
we have 3 cases to handle. Either the 'skip' case which skips the
completion of the job. The 'wait' case which was the default before and
still is when $completion is undefined. And the new 'wait_noswap' case
which is used for the live migration.
If 'wait_noswap' is specified, we issue a 'block-job-cancel' once the block
job is in 'ready' state. This completes the block job without swapping the
disks.
clone_disk always uses 'block-job-cancel' via the qemu_blockjobs_cancel
sub.
Signed-off-by: Mira Limbeck <m.limbeck@proxmox.com>
removes safe_string_ne and safe_num_ne code which is now shared in
GuestHelpers. also change all the calls to use the shared definitions.
Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
The initialization for the drive keys in $confdesc is changed
to be a single for-loop iterating over the keys of $drivedesc_hash and
the initialization of the unusedN keys is move to directly below it.
To avoid the need to change all the call sites, functions with more than
a few callers are exported from the submodule and imported into QemuServer.pm.
For callers of the now imported functions within QemuServer.pm, the prefix
PVE::QemuServer is dropped, because it is unnecessary and now even confusing.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
which contains the full descriptions of the drives, and
make parse_drive not depend on $confdesc anymore.
In preparation to moving drive-related code to its own module.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Allow a user to add a virtio-rng-pci (an emulated hardware random
number generator) to a VM with the rng0 setting. The setting is
version_guard()-ed.
Limit the selection of entropy source to one of three:
/dev/urandom (preferred): Non-blocking kernel entropy source
/dev/random: Blocking kernel source
/dev/hwrng: Hardware RNG on the host for passthrough
QEMU itself defaults to /dev/urandom (or the equivalent getrandom()
call) if no source file is given, but I don't fully trust that
behaviour to stay constant, considering the documentation [0] already
disagrees with the code [1], so let's always specify the file ourselves.
/dev/urandom is preferred, since it prevents host entropy starvation.
The quality of randomness is still good enough to emulate a hwrng, since
a) it's still seeded from the kernel's true entropy pool periodically
and b) it's mixed with true entropy in the guest as well.
Additionally, all sources about entropy predicition attacks I could find
mention that to predict /dev/urandom results, /dev/random has to be
accessed or manipulated in one way or the other - this is not possible
from a VM however, as the entropy we're talking about comes from the
*hosts* blocking pool.
More about the entropy and security implications of the non-blocking
interface in [2] and [3].
Note further that only one /dev/hwrng exists at any given time, if
multiple RNGs are available, only the one selected in
'/sys/devices/virtual/misc/hw_random/rng_current' will feed the file.
Selecting this is left as an exercise to the user, if at all required.
We limit the available entropy to 1 KiB/s by default, but allow the user
to override this. Interesting to note is that the limiter does not work
linearly, i.e. max_bytes=1024/period=1000 means that up to 1 KiB of data
becomes available on a 1000 millisecond timer, not that 1 KiB is
streamed to the guest over the course of one second - hence the
configurable period.
The default used here is the same as given in the QEMU documentation [0]
and has been verified to affect entropy availability in a guest by
measuring /dev/random throughput. 1 KiB/s is enough to avoid any
early-boot entropy shortages, and already has a significant impact on
/dev/random availability in the guest.
[0] https://wiki.qemu.org/Features/VirtIORNG
[1] https://git.qemu.org/?p=qemu.git;a=blob;f=crypto/random-platform.c;h=f92f96987d7d262047c7604b169a7fdf11236107;hb=HEAD
[2] https://lwn.net/Articles/261804/
[3] https://lwn.net/Articles/808575/
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
1. Avoids the error
"VM 111 qmp command 'block_resize' failed - The new size must be a multiple of 512"
for qcow2 disks.
2. Because volume_import expects disk sizes to be a multiple of 1 KiB.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
Some of the recent QMP changes require at least 2.8.0, but since the
oldest version we officially package for 6.x is 4.0.0 anyway, checking
for at least 3.0 should not break anyone's setup.
Note that this does not affect machine version checks, only the
installed QEMU binary version.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
Live-migrating a VM with more than 14 SCSI disks to a node that doesn't
support it yet is broken. Use a bumped pve-version to represent that and
give the user a nice error message instead.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
The previously introduced approach can fail for pinned versions when a
new QEMU release is introduced. The saner approach is to use a mapping
that gives one pve-version for each QEMU release.
Fortunately, the old system has not been bumped yet, so we can still
change it without too much effort.
QEMU versions without a mapping are assumed to be pve0, 4.1 is mapped to
pve1 since thats what we had as our default previously.
Pinned machine versions (i.e. pc-i440fx-4.1) are always assumed to be
pve0, for specific pve-versions they'd have to be pinned as well (i.e.
pc-i440fx-4.1+pve1).
The new logic also makes the pve-version dynamic, and starts VMs with
the lowest possible 'feature-level', i.e. if a feature is only available
with 4.1+pve2, but the VM isn't using it, we still start it with
4.1+pve0.
We die if we don't support a version that is requested from us. This
allows us to use the pve-version as live-migration blocks (i.e. bumping
the version and then live-migrating a VM which uses the new feature (so
is running with the bumped version) to an outdated node will present the
user with a helpful error message and fail instead of silently modifying
the config and only failing *after* the migration).
$version_guard is introduced in config_to_command to use for features
that need to check pve-version, it automatically handles selecting the
newest necessary pve-version for the VM.
Tests have to be adjusted, since all of them now resolve to pve0 instead
of pve1. EXPECT_ERROR matching is changed to use 'eq' instead of regex
to allow special characters in error messages.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
query-cpus has been deprecated since 2.12.0 [0] in favor of
query-cpus-fast, which no longer incurs a guest performance penalty on
the guest. The returned information is the same as far as our use case
is concerned.
[0] https://qemu.weilnetz.de/doc/qemu-doc.html#Deprecated-features
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
'device' is deprecated since 2.8 in favor of 'id' [0], but since we
always consistently set the id on our drives anyway we can substitute it
easily.
[0] see files qapi/block.json and qapi/block-core.json in QEMU source
code, the online documentation doesn't mention it AFAICT
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
...and cleanup surrounding code a bit.
'change' is deprecated, and according to the qapi definition in QEMU it
is 'strongly recommended' to avoid using it.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
to achieve this we have to add 3 new scsihw addresses since lsi
controllers can only hold 7 scsi drives
we go up to 31, since this is the limit for virtio-scsi-single devices
we have reserved (we can increase this in the future)
to make it more future proof, we add a new pci bridge under pci
bridge 1, so we have to adapt the bridge adding code (we did not
need this for q35 previously)
impact on live migration:
since on older versions of qemu-server we do not have those config
settings, there is no problem from old -> new
new->old is not supported anyway and this breaks so that
the vm crashes and loses the configs for scsi15-30
(same behaviour as e.g. with audio0 and migration from new->old)
tested with 31 scsi disk on
i440fx + virtio-scsi
i440fx + lsi
q35 + virtio-scsi
q35 + lsi
with ovmf + seabios
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
and adapt the tests
this does not impact live migration, since the order here does not
change the device layout
we want this to consistently have the readconfig first
Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
from hotplug_pending we go into 'vmconfig_update_disk', where we check the
hotpluggability of options.
add 'ssd' there as a non-hotpluggable option (since we'd have to unplug/plug to
change the drive type)
Signed-off-by: Oguz Bektas <o.bektas@proxmox.com>
The package will be used for custom CPU models as a SectionConfig, hence
the name. For now we simply move some CPU related helper functions and
declarations over from QemuServer to reduce clutter there.
Exports are to avoid changing all call sites, functions have useful
names on their own.
Signed-off-by: Stefan Reiter <s.reiter@proxmox.com>
As 'qemu_img_format' just matches a regex, this doesn't make much of
a difference, but AFAICT all other calls of 'qemu_img_format' use 'volname'.
Signed-off-by: Fabian Ebner <f.ebner@proxmox.com>
