Am 19.06.23 um 14:25 schrieb Alexandre Derumier:
> Signed-off-by: Alexandre Derumier
> ---
> PVE/QemuServer/CPUConfig.pm | 15 +++
> 1 file changed, 15 insertions(+)
>
> diff --git a/PVE/QemuServer/CPUConfig.pm b/PVE/QemuServer/CPUConfig.pm
> index fb0861b..cfccbd4 100644
> --- a/PVE
Signed-off-by: Fiona Ebner
---
Depends on https://lists.proxmox.com/pipermail/pve-devel/2023-June/057635.html
Used a new window, because it's small and couldn't find a good
fit with the existing ones. Maybe SafeDestroy, but not in name and
would require a few modifications too.
Not sure about u
by removing the confusing buttons in the toolbar and adding them as
actions in an actioncolumn. There a only relevant actions are visible
and get a more expressive tooltip
with this, we now differentiate between 4 modes of the edit window:
* create a new mapping altogether
- shows all fields
* e
they got lost in my last rebase/refactor.
the onLoadCallBack is used to check by the window if there are iommu
groups at all, and the checkIsolated function checks if the selected
ones are in a separate group (in regards to the other devices)
Signed-off-by: Dominik Csapak
---
www/manager6/form/
by changing into 'mapping matches host data' which indicates that the
configured values matches the host information
also for the pci and usb map selectors
Signed-off-by: Dominik Csapak
---
www/manager6/form/PCIMapSelector.js | 2 +-
www/manager6/form/USBMapSelector.js | 2 +-
www/manager6/tr
by disallowing nodes to be selected where a mapping already exists
and not preselecting a node
Signed-off-by: Dominik Csapak
---
www/manager6/window/PCIMapEdit.js | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/www/manager6/window/PCIMapEdit.js
b/www/manager6/windo
Signed-off-by: Alexandre Derumier
---
PVE/QemuServer/CPUConfig.pm | 15 +++
1 file changed, 15 insertions(+)
diff --git a/PVE/QemuServer/CPUConfig.pm b/PVE/QemuServer/CPUConfig.pm
index fb0861b..cfccbd4 100644
--- a/PVE/QemuServer/CPUConfig.pm
+++ b/PVE/QemuServer/CPUConfig.pm
@@ -66
Am 19.06.23 um 11:29 schrieb Aaron Lauterer:
>
> @@ -4876,11 +4876,13 @@ sub foreach_volid {
> $volhash->{$volid}->{shared} = 1 if $drive->{shared};
>
> $volhash->{$volid}->{referenced_in_config} //= 0;
> - $volhash->{$volid}->{referenced_in_config} = 1 if !defined($snapname);
>
Am 19.06.23 um 11:29 schrieb Aaron Lauterer:
> This patch series changes the behavior during guest migrations:
>
> Don't scan all storages for potential images belonging to the guest.
> Only migrate images referenced in the config.
> This made it necessary to handle pending changes explicitly whic
Am 19.06.23 um 11:29 schrieb Aaron Lauterer:
> @@ -277,6 +280,12 @@ sub phase1 {
> # finally all current volumes
> PVE::LXC::Config->foreach_volume_full($conf, { include_unused => 1 },
> $test_mp);
>
> +for my $path (keys %$path_to_volid) {
> + my @volids = keys $path_to_volid-
Am 19.06.23 um 11:29 schrieb Aaron Lauterer:
> @@ -427,6 +429,12 @@ sub scan_local_volumes {
> }
> });
>
> + for my $path (keys %$path_to_volid) {
> + my @volids = keys $path_to_volid->{$path}->%*;
> + die "detected not supported aliased volumes: '" . join("
Am 19/06/2023 um 11:13 schrieb Dominik Csapak:
> the 'errors' property changed to 'checks' so we have to adapt here too
>
> Signed-off-by: Dominik Csapak
> ---
> www/manager6/form/PCIMapSelector.js | 14 +++---
> 1 file changed, 7 insertions(+), 7 deletions(-)
>
>
applied series, thank
ping
On 1/18/23 14:57, Markus Frank wrote:
qemu-server:
v5:
* set $kvm to 1 if is_native, so that api kvm check works.
v4:
* added kvm/q35 checks in API
* reused pve-qemu-machine
v3:
* replaced old machine type with property-string with viommu-parameter
v2:
* moved viommu-parameter inside of
Am 19.06.23 um 11:19 schrieb DERUMIER, Alexandre:
> Le lundi 19 juin 2023 à 10:46 +0200, Fiona Ebner a écrit :
>
>> I also noticed that
>> Denverton{-v2,-v3}
>> Snowridge{-v2,-v3,-v4}
>
> Both are atom cpu. I don't known if we want to support them ? (they
> seem to support vtx, so it should work
Am 19.06.23 um 10:38 schrieb Alexandre Derumier:
> add doc for differents cpu models including
> new x86-64-vX models
>
> changelog v2: x86-64-v2 is compatible with >= opteron_g3
>
> Signed-off-by: Alexandre Derumier
> ---
> qm.adoc | 145 +---
When scanning all configured storages for volumes belonging to the
container, the migration could easily fail if a storage is not
available, but enabled. That storage might not even be used by the
container at all.
By not doing that and only looking at the disk images referenced in the
config, we
Reviewed-by: Fiona Ebner
Signed-off-by: Aaron Lauterer
---
changes since v4: none
test/run_qemu_migrate_tests.pl | 75 ++
1 file changed, 75 insertions(+)
diff --git a/test/run_qemu_migrate_tests.pl b/test/run_qemu_migrate_tests.pl
index 3d5eb8d..4373a38 100755
Aliased volumes (referencing the same volume multiple times) can lead to
unexpected behavior in a migration.
Therefore, stop the migration in such a case.
The check works by comparing the path returned by the storage plugin.
This means that we should be able to catch the common situations where
i
Aliased volids can lead to unexpected behavior in a migration.
An aliased volid can happen if we have two storage configurations,
pointing to the same place. The resulting 'path' for a disk image
will be the same.
Therefore, stop the migration in such a case.
The check works by comparing the path
Make it possible to optionally iterate over disks in the pending section
of VMs, similar as to how snapshots are handled already.
This is for example useful in the migration if we don't want to rely on
the scanning of all storages.
All calling sites are adapted and enable it, except for
QemuConfi
This patch series changes the behavior during guest migrations:
Don't scan all storages for potential images belonging to the guest.
Only migrate images referenced in the config.
This made it necessary to handle pending changes explicitly which had
been covered by the storage scan.
We also added
Reviewed-by: Fiona Ebner
Signed-off-by: Aaron Lauterer
---
changes since v4: none
pvesm.adoc | 6 ++
1 file changed, 6 insertions(+)
diff --git a/pvesm.adoc b/pvesm.adoc
index d250e0b..49eb972 100644
--- a/pvesm.adoc
+++ b/pvesm.adoc
@@ -177,6 +177,12 @@ zfspool: local-zfs
content
Since we don't scan all storages for matching disk images anymore for a
migration we don't have any images found via storage alone. They will be
referenced in the config somewhere.
Therefore, there is no need for the 'storage' ref.
The 'referenced_in_config' is not really needed and can apply to b
When scanning all configured storages for disk images belonging to the
VM, the migration could easily fail if a storage is not available, but
enabled. That storage might not even be used by the VM at all.
By not scanning all storages and only looking at the disk images
referenced in the VM config,
Reviewed-by: Fiona Ebner
Signed-off-by: Aaron Lauterer
---
changes since v4: none
test/run_qemu_migrate_tests.pl | 64 ++
1 file changed, 64 insertions(+)
diff --git a/test/run_qemu_migrate_tests.pl b/test/run_qemu_migrate_tests.pl
index fedbc32..3d5eb8d 100755
-
All calling sites except for QemuConfig.pm::get_replicatable_volumes()
already enabled it. Making it the non-configurable default results in a
change in the VM replication. Now a disk image only referenced in the
pending section will also be replicated.
Signed-off-by: Aaron Lauterer
---
changes
Le lundi 19 juin 2023 à 10:46 +0200, Fiona Ebner a écrit :
> Am 09.06.23 um 12:27 schrieb Alexandre Derumier:
> > Signed-off-by: Alexandre Derumier
> > ---
> > PVE/QemuServer/CPUConfig.pm | 14 ++
> > 1 file changed, 14 insertions(+)
> >
> > diff --git a/PVE/QemuServer/CPUConfig.pm
>
and mdev change.
giving an empty array did not have the desired effect of resetting the
selection, but giving no parameter at all does.
this now also clears the selection when the mdev filter/config changed
(was just forgotten)
Signed-off-by: Dominik Csapak
---
www/manager6/form/MultiPCISelect
when there is more than one function for a device, indent the individual
functions. This sets them visually apart from the 'pass all through as
one' entry
We have to use a html entity here, as extjs trims the normal whitespace.
Signed-off-by: Dominik Csapak
---
www/manager6/form/MultiPCISelecto
by
* moving the node to the left column and changing the label
* moving the mdev filter to the right column
* show only the create button for new node mappings (otherwise we'd have
a reset button here that cannot do anything useful)
Signed-off-by: Dominik Csapak
---
www/manager6/window/PCIMapE
the 'errors' property changed to 'checks' so we have to adapt here too
Signed-off-by: Dominik Csapak
---
www/manager6/form/PCIMapSelector.js | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
diff --git a/www/manager6/form/PCIMapSelector.js
b/www/manager6/form/PCIMapSelector.
aka a 'splitter'. that way the user can determine how much of each panel
he wants to see himself
Signed-off-by: Dominik Csapak
---
www/manager6/dc/Config.js | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/www/manager6/dc/Config.js b/www/manager6/dc/Config.js
index
Am 09.06.23 um 12:27 schrieb Alexandre Derumier:
> Signed-off-by: Alexandre Derumier
> ---
> PVE/QemuServer/CPUConfig.pm | 14 ++
> 1 file changed, 14 insertions(+)
>
> diff --git a/PVE/QemuServer/CPUConfig.pm b/PVE/QemuServer/CPUConfig.pm
> index fb0861b..77fc682 100644
> --- a/PVE/
add doc for differents cpu models including
new x86-64-vX models
changelog v2: x86-64-v2 is compatible with >= opteron_g3
Signed-off-by: Alexandre Derumier
---
qm.adoc | 145 +---
1 file changed, 138 insertions(+), 7 deletions(-)
diff --git a
Am 16/06/2023 um 12:57 schrieb Maximiliano Sandoval:
> We make sure that the Glib main loop is closed before exiting the application.
>
> Ideally all of this, among other things, is handled automatically when using
> Gtk3::ApplicationWindow and Glib::IO::Application->run(), but that needs
> https:
Am 15/06/2023 um 14:14 schrieb Maximiliano Sandoval:
> Small series of patches to drop the use of properties and methods that were
> deprecated in GTK 3.
>
> Maximiliano Sandoval (4):
> stop using Gtk3::Misc->set_alignment
> stop using Gtk3::HBox and Gtk3::VBox
> stop using Gtk3::HSeparator
a 4GiB static memory is needed for DMA+boot memory, as this memory
is almost always un-unpluggeable.
1 virtio-mem pci device is setup for each numa node on pci.4 bridge
virtio-mem use a fixed blocksize with 32000 blocks
Blocksize is computed from the maxmemory-4096/32000 with a minimum of
2MiB to
Signed-off-by: Alexandre Derumier
---
test/cfg2cmd/memory-virtio-hugepages-1G.conf | 12 +++
.../memory-virtio-hugepages-1G.conf.cmd | 35 +++
test/cfg2cmd/memory-virtio-max.conf | 11 ++
test/cfg2cmd/memory-virtio-max.conf.cmd | 35 +
If some memory can be removed on a specific node,
we try to rebalance again on other nodes
Signed-off-by: Alexandre Derumier
---
PVE/QemuServer/Memory.pm | 59 +---
1 file changed, 43 insertions(+), 16 deletions(-)
diff --git a/PVE/QemuServer/Memory.pm b/PVE/
default kernel vhost config only support 64 slots by default,
for performance since 2015.
Original memory hotplug code was done before, using qemu
max supported 255 slots.
To reach max mem (4TB), we used incremental dimm size.
Instead of dynamic memory size, use 1 static dimm size, compute
from
Signed-off-by: Alexandre Derumier
---
PVE/QemuServer/Memory.pm | 12 +---
1 file changed, 5 insertions(+), 7 deletions(-)
diff --git a/PVE/QemuServer/Memory.pm b/PVE/QemuServer/Memory.pm
index 12e6ee4..ee73a0a 100644
--- a/PVE/QemuServer/Memory.pm
+++ b/PVE/QemuServer/Memory.pm
@@ -151,1
Signed-off-by: Alexandre Derumier
---
www/manager6/qemu/MemoryEdit.js | 52 +++--
1 file changed, 43 insertions(+), 9 deletions(-)
diff --git a/www/manager6/qemu/MemoryEdit.js b/www/manager6/qemu/MemoryEdit.js
index 5e91dc9b..be7903a2 100644
--- a/www/manager6/qemu/Me
This patch series rework the current memory hotplug + virtiomem.
memory option now have extra options:
memory: [[current=]] [,max=] [,virtio=<1|0>]
ex: memory: current=1024,max=131072,virtio=1
for classic memory hotplug, when maxmemory is defined,
we use 64 fixed size dimm.
The max option is a
Signed-off-by: Alexandre Derumier
---
www/manager6/qemu/HardwareView.js | 17 +++--
1 file changed, 15 insertions(+), 2 deletions(-)
diff --git a/www/manager6/qemu/HardwareView.js
b/www/manager6/qemu/HardwareView.js
index af35a980..0b17ef0d 100644
--- a/www/manager6/qemu/HardwareVie
Signed-off-by: Alexandre Derumier
---
PVE/API2/Qemu.pm | 7 ++--
PVE/QemuConfig.pm | 4 +--
PVE/QemuMigrate.pm| 6 ++--
PVE/QemuServer.pm | 27
PVE/QemuServer/Helpers.pm | 3 +-
PVE/QemuServer/Memory.pm | 67 --
Signed-off-by: Alexandre Derumier
---
test/cfg2cmd/memory-max-128G.conf | 11
test/cfg2cmd/memory-max-128G.conf.cmd | 86 +++
test/cfg2cmd/memory-max-512G.conf | 11
test/cfg2cmd/memory-max-512G.conf.cmd | 58 ++
4 files changed, 166 inser
verify than defined vm memorymax is not bigger than
host cpu supported memory
Add add early check in update vm api
Signed-off-by: Alexandre Derumier
---
PVE/API2/Qemu.pm | 35 +--
PVE/QemuServer/Memory.pm | 19 ++-
2 files changed, 43 inse
max can be multiple of 64GiB only,
The dimm size is compute from the max memory
we can have 64 slots:
64GiB = 64 slots x 1GiB
128GiB = 64 slots x 2GiB
..
4TiB = 64 slots x 64GiB
Also, with numa, we need to share slot between (up to 8) sockets.
64 is a multiple of 8,
64GiB = 8 sockets * 8 slots
Signed-off-by: Alexandre Derumier
---
PVE/QemuServer/Memory.pm | 37 +++--
1 file changed, 19 insertions(+), 18 deletions(-)
diff --git a/PVE/QemuServer/Memory.pm b/PVE/QemuServer/Memory.pm
index aaa94cc..12e6ee4 100644
--- a/PVE/QemuServer/Memory.pm
+++ b/PVE/Qem
Am 16/06/2023 um 15:05 schrieb Dominik Csapak:
> pve-manager:
>
> Dominik Csapak (14):
> api: add resource map api endpoints for PCI and USB
> ui: parser: add helper for lists of property strings
> ui: form/USBSelector: make it more flexible with nodename
> ui: form: add PCIMapSelector
>
Le lundi 19 juin 2023 à 09:11 +0200, Thomas Lamprecht a écrit :
> Am 08/06/2023 um 11:51 schrieb Alexandre Derumier:
> > Here the new builtin models:
> >
> >
> > x86-64-v1 : not implemented, as it's basicaly qemu64|kvm64 -vme,-
> > cx16 for compat Opteron_G1 from 2004
> > so will use
Am 16/06/2023 um 15:05 schrieb Dominik Csapak:
> explain why someone would want it, how to configure and which privileges
> are necessary
>
> Signed-off-by: Dominik Csapak
> ---
> changes from v6:
> * added small note about only one usb device per node per map
>
> qm-pci-passthrough.adoc | 8 +
Am 08/06/2023 um 11:51 schrieb Alexandre Derumier:
> Here the new builtin models:
>
>
> x86-64-v1 : not implemented, as it's basicaly qemu64|kvm64 -vme,-cx16 for
> compat Opteron_G1 from 2004
> so will use it as qemu64|kvm64 is higher are not working on
> opteron_g1 anyway
>
> x86-
53 matches
Mail list logo
