Signed-off-by: Aaron Lauterer
---
nothing has changed to previous series
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/src/PVE/LXC/Config.pm b/src/PVE/LXC/Config.pm
index 8d1854a..9fab175 100644
--- a/src/PVE/LXC/Config.pm
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v6 -> v7:
* change key `data` to `volume_config` in return hash f
with the actual implementations.
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-June/044000.html
[1] https://pve.proxmox.com/pipermail/pve-devel/2020-June/044001.html
[2] https://pve.proxmox.com/pipermail/pve-devel/2020-June/044009.html
qemu-server: Aaron Lauterer (1):
vzdump: move include
as `data` was a bit too generic we now use `volume_config` in the actual
implementations. Thus we should adapt the description as well.
Tab spacing for the other keys has been adapted for easier readabilty.
Signed-off-by: Aaron Lauterer
---
turns out that the `+` in the git diff pushes
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v6 -> v7: incorporate suggestions, mainly
* change key `d
ping in case this was drowned by all the other patches sent :)
On 5/20/20 4:05 PM, Aaron Lauterer wrote:
The documentation only covered replacing the disk with systemd-boot but
not if grub is used.
Signed-off-by: Aaron Lauterer
---
Once this is applied and rolled out (in the Wiki), we should
a few small nits and some suggestions on the phrasing inline
On 6/18/20 11:05 AM, Stefan Reiter wrote:
Signed-off-by: Stefan Reiter
---
Requires qemu-server with patch 2 applied.
Makefile | 1 +
cpu-models.conf.adoc | 82 +++
On 6/17/20 3:50 PM, Thomas Lamprecht wrote:
Am 6/17/20 um 2:13 PM schrieb Aaron Lauterer:
The `guest include` logic handling `all` and `exclude` parameters was in
the `PVE::VZDump->exec_backup()` method. Moving this logic into the
`get_included_guests` method allows us to simpl
Signed-off-by: Aaron Lauterer
---
nothing changed to previous series
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/src/PVE/LXC/Config.pm b/src/PVE/LXC/Config.pm
index 8d1854a..9fab175 100644
--- a/src/PVE/LXC/Config.pm
+++ b
Now, with the logic for `all` and `exclude` in the same single method,
additional tests for these cases are possible.
Adapt to hash return value of `get_included_guests`.
Signed-off-by: Aaron Lauterer
---
v5 -> v6: adapt call to PVE::VZDump::get_included_guests()
t
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v5 -> v6: create hash with volume data directly in push call inst
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v5 -> v6: create hash with volume data directly in push call inst
://pve.proxmox.com/pipermail/pve-devel/2020-March/042420.htmli
manager: Aaron Lauterer (2):
vzdump: move remaining guest include logic to single method
vzdump: test: adapt and add more tests that are possible now
PVE/API2/VZDump.pm | 19 +++--
PVE/VZDump.pm | 80
cks are kept where they are to be able to handle missing
permissions according to the current context. The old behavior to die
on a backup job when the user is missing the permission to a guest and
the job is not an 'all' or 'exclude' job is kept.
Signed-off-by: Aaron Lauterer
---
v5 -&
On 6/5/20 8:39 PM, Thomas Lamprecht wrote:
On 5/6/20 11:57 AM, Aaron Lauterer wrote:
[...]
@@ -1169,34 +1158,39 @@ sub get_included_guests {
my $nodename = PVE::INotify::nodename();
my $vmids = [];
+my $vmids_per_node = {};
+
+my $vmlist = PVE::Cluster::get_vmlist
As a first step to make the whole guest include logic more testable the
part from the API endpoint has been moved to its own method with as
little changes as possible.
Everything concerning `all` and `exclude` logic is still in the
PVE::VZDump->exec_backup() method.
Signed-off-by: Aaron Laute
Signed-off-by: Aaron Lauterer
---
v2 -> v3:
* define pool in separate hash instead of in the return value
* changed the adding and processing of tests as suggested by Thomas [0]
v1 -> v2: adapt handling of return values, closer to what is used in
production code.
[0] https://pve.proxm
On 6/5/20 8:43 PM, Thomas Lamprecht wrote:
On 5/6/20 11:57 AM, Aaron Lauterer wrote:
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron
On 6/5/20 8:39 PM, Thomas Lamprecht wrote:
On 5/6/20 11:57 AM, Aaron Lauterer wrote:
[...]
diff --git a/PVE/API2/VZDump.pm b/PVE/API2/VZDump.pm
index 68a3de89..77dac1b1 100644
--- a/PVE/API2/VZDump.pm
+++ b/PVE/API2/VZDump.pm
@@ -69,19 +69,26 @@ __PACKAGE__->register_met
Besides sending a new version for this series, I will also send a new
version of series with the actual changes to the backup logic as the
hints given here also apply there.
On 6/5/20 8:57 AM, Thomas Lamprecht wrote:
On 5/4/20 4:08 PM, Aaron Lauterer wrote:
Signed-off-by: Aaron Lauterer
On 6/3/20 4:35 PM, Thomas Lamprecht wrote:
On 5/4/20 4:08 PM, Aaron Lauterer wrote:
[...]
+sub get_included_guests {
+my ($self, $job) = @_;
do we need $self here? Why not call it like:
PVE::VZDump::get_included_guests($params) ?
You're right. No need for it AFAICT
+
+my
The documentation only covered replacing the disk with systemd-boot but
not if grub is used.
Signed-off-by: Aaron Lauterer
---
Once this is applied and rolled out (in the Wiki), we should remove the
old steps in the ZFS tips and tricks page [0] and link to this one.
[0]
https
need more consideration, or better, a reboot.
Signed-off-by: Aaron Lauterer
---
images/screenshot/boot-grub.png| Bin 0 -> 14810 bytes
images/screenshot/boot-systemdboot.png | Bin 0 -> 4527 bytes
system-booting.adoc| 39 +
3 files chang
Show the real path of the partition in case when the basename couldn't
be determined and the partition given is a symlinked one like
/dev/disk/by-id//
Signed-off-by: Aaron Lauterer
---
I hope this is not a too hacky solution :)
bin/pve-efiboot-tool | 5 -
1 file changed, 4 insertions
. It is
stored in `bdev`.
Signed-off-by: Aaron Lauterer
---
bin/pve-efiboot-tool | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/bin/pve-efiboot-tool b/bin/pve-efiboot-tool
index eab01d7..d54b443 100755
--- a/bin/pve-efiboot-tool
+++ b/bin/pve-efiboot-tool
@@ -96,7 +96,7 @@ format
Tested-By: Aaron Lauterer
On 5/18/20 9:37 AM, Fabian Ebner wrote:
Signed-off-by: Fabian Ebner
---
Changes from v1:
* die/warn depending on force (thanks to Thomas and Aaron for the
suggestion)
* don't die/warn if VM is not replicated at all
PVE/API2/Qemu.pm | 13
On 5/14/20 3:11 PM, Thomas Lamprecht wrote:
On 5/14/20 3:09 PM, Aaron Lauterer wrote:
Works as expected.
Throws a warning when migrating a running VM to a node to which no replication
job is set up.
Since I wasn't following all of the previous off-list discusion: Do we just
want
Works as expected.
Throws a warning when migrating a running VM to a node to which no
replication job is set up.
Since I wasn't following all of the previous off-list discusion: Do we
just want a warning that could be overseen quite easily instead of dying?
Nonetheless:
Tested-By: Aaron
Signed-off-by: Aaron Lauterer
---
pvecm.adoc | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/pvecm.adoc b/pvecm.adoc
index 2866582..2cdb92b 100644
--- a/pvecm.adoc
+++ b/pvecm.adoc
@@ -61,7 +61,7 @@ Requirements
* Date and time have to be synchronized.
-* SSH tunnel
The network config examples use the underscore but our tooling
generates the configs with dashes.
Signed-off-by: Aaron Lauterer
---
pve-network.adoc | 80
pvecm.adoc | 6 ++--
2 files changed, 43 insertions(+), 43 deletions(-)
diff --git
With one optional nit inline:
Reviewed-By: Aaron Lauterer
On 5/7/20 4:16 PM, Alwin Antreich wrote:
as a short description about the different algorithms in use by vzdump.
Signed-off-by: Alwin Antreich
---
v1 -> v2:
* incorporate Aaron's suggestions
https://pve.proxmox.
Some nits inline.
either in s/x/y/ syntax or just a rewrite of a sentence.
On 5/7/20 2:42 PM, Alwin Antreich wrote:
as a short description about the different algorithms in use by vzdump.
Signed-off-by: Alwin Antreich
---
vzdump.adoc | 36 +++-
1 file
This patch adds a new API endpoint that returns a list of included
guests, their volumes and whether they are included in a backup.
The output is formatted to be used with the extJS tree panel.
Signed-off-by: Aaron Lauterer
---
The return types are `qemu`, `lxc` and `unknown`. The latter
Adds a new `Summary` button to the backup panel which opens a new
window containing a searchable view of all guests and whether they are
backed up by any job or not.
Signed-off-by: Aaron Lauterer
---
I took the code for the search field from the ResourceGrid, simplified
it and adapted it more
The new detail view for backup jobs shows the settings similar to the
edit dialog but read only. Additionally it does show a list of all
included guests with their volumes and whether these volumes will be
included in the backup.
Signed-off-by: Aaron Lauterer
---
v2->v3:
* made the comment
beforehand to work correctly.
Changes are mostly incorporating feedback and making the code nicer.
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-May/043340.html
[1] https://pve.proxmox.com/pipermail/pve-devel/2020-May/043399.html
Aaron Lauterer (5):
api: backup: add endpoint to list included
Moving the following renderers to Utils.js to be able to use them in
more than one place:
* render_backup_days_of_week
* render_backup_selection
Signed-off-by: Aaron Lauterer
---
v1 -> v2 ->v3: rebased
www/manager6/Utils.js | 61 +++
www/manag
is the `included_status` which
returns a list of all guests and wheter they are included in a backup
job or not backed up by any.
The top level index endpoint is only a small stub for now to be available for
a more generic summary endpoint in the future.
Signed-off-by: Aaron Lauterer
---
v1->
fixes behavior introduced with commit
940e2a3a06b3ea47aae144519e2aaa881a80e437
Qemu 4.1 will fail to start a guest with an audio device set with
`Property '.audiodev' not found`.
Signed-off-by: Aaron Lauterer
---
Users reported this problem in the forum today [0].
I was able to reproduce
I forgot to mention in the notes, that this patch and the other manager
patch with the tests needs the following series applied beforehand!
https://pve.proxmox.com/pipermail/pve-devel/2020-May/043340.html
___
pve-devel mailing list
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v4->v5:
* use new `foreach_volume`
* change $ret_volu
Signed-off-by: Aaron Lauterer
---
nothing changed to previous series'
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/src/PVE/LXC/Config.pm b/src/PVE/LXC/Config.pm
index dcc8755..34f657a 100644
--- a/src/PVE/LXC/Config.pm
+++ b
cks are kept where they are to be able to handle missing
permissions according to the current context. The old behavior to die
on a backup job when the user is missing the permission to a guest and
the job is not an 'all' or 'exclude' job is kept.
Signed-off-by: Aaron Lauterer
---
v4 -> v5:
Now, with the logic for `all` and `exclude` in the same single method,
additional tests for these cases are possible.
Adapt to hash return value of `get_included_guests`.
Signed-off-by: Aaron Lauterer
---
test/vzdump_guest_included_test.pl | 123 ++---
1 file changed
call in QemuConfig.pm
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-May/043340.html
[1] https://pve.proxmox.com/pipermail/pve-devel/2020-April/042753.html
[5] https://pve.proxmox.com/pipermail/pve-devel/2020-March/042420.html
manager: Aaron Lauterer (2):
vzdump: move remaining guest include
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v4->v5:
* use new foreach_volume call
* change $ret_volu
As a first step to make the whole guest include logic more testable the
part from the API endpoint has been moved to its own method with as
little changes as possible.
Everything concerning `all` and `exclude` logic is still in the
PVE::VZDump->exec_backup() method.
Signed-off-by: Aaron Laute
Signed-off-by: Aaron Lauterer
---
v1 -> v2: adapt handling of return values, closer to what is used in
production code.
test/Makefile | 5 +-
test/vzdump_guest_included_test.pl | 191 +
2 files changed, 195 insertions(+), 1 deletion(-)
cre
Sorry, I didn't get to read v1 in time. Some nits regarding grammar,
reading flow and such inline:
Other than those:
Reviewed-By: Aaron Lauterer
On 4/30/20 2:27 PM, Dominik Csapak wrote:
explaining the main Requirements and limitations, as well as the
most important sync options
Signed-off
Signed-off-by: Aaron Lauterer
---
I forgot the makefile, please ignore the other patch and use this one.
test/Makefile | 5 +-
test/vzdump_guest_included_test.pl | 190 +
2 files changed, 194 insertions(+), 1 deletion(-)
create mode 100755
Signed-off-by: Aaron Lauterer
---
test/vzdump_guest_included_test.pl | 190 +
1 file changed, 190 insertions(+)
create mode 100755 test/vzdump_guest_included_test.pl
diff --git a/test/vzdump_guest_included_test.pl
b/test/vzdump_guest_included_test.pl
new file mode
kup() method.
Signed-off-by: Aaron Lauterer
---
As talked with @thomas on[0] and off list, this patch series is meant to
have more confidence in the ongoing changes.
My other ongoing patch series [1] will move the all the logic, even the
one in the `exec_backup()` method into one single method.
Any news here? I'm asking because I just saw the deprecation warning in
some testing :)
On 3/27/20 1:24 PM, Aaron Lauterer wrote:
On 3/27/20 12:22 PM, Thomas Lamprecht wrote:
On 3/16/20 11:02 AM, Aaron Lauterer wrote:
With Qemu 4.2 a new `audiodev` property was introduced [0] to explicitly
On 4/7/20 5:55 PM, Thomas Lamprecht wrote:
> On 4/6/20 4:24 PM, Aaron Lauterer wrote:
>> [...]
>> +description => 'Root node of the tree object. Children
represent guests, grandchildren represent volumes of that guest.',
>> +properties => {
>
ZFS supports the -p flag in the list command since a few years now.
Let us use the real byte values and avoid the error prone calculation
from human readable numbers that can lead to incorrect numbers if the
reported human readable value is a rounded number.
Signed-off-by: Aaron Lauterer
Signed-off-by: Aaron Lauterer
---
this was the only other place I found where the use of the -p flag might
be useful
PVE/Storage/ZFSPoolPlugin.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/PVE/Storage/ZFSPoolPlugin.pm b/PVE/Storage/ZFSPoolPlugin.pm
index ad2fe9b
there is no calculation performed and
thus we need to explicitly cast the size to an integer before returning
it.
Signed-off-by: Aaron Lauterer
---
v1->v2: made the -p flag more obvious in the commit msg by adding a dash
Removal of zfs_parse_size and using the -p flag where appropiate is d
'-o', 'refquota', "$scfg->{pool}/$basename");
1 chomp($size);
I don't see much chances for a bug here as the value is used right away
in the next step but it would definitely be more precise to use the -p flag.
On April 3, 2020 2:29 pm, Aaron Lauterer wrote:
Getting the volu
is the `included_status` which
returns a list of all guests and wheter they are included in a backup
job or not backed up by any.
The top level index endpoint is left unsused for now to be available for
a more generic summary endpoint in the future.
Signed-off-by: Aaron Lauterer
---
Feedback regarding
Moving the following renderers to Utils.js to be able to use them in
more than one place:
* render_backup_days_of_week
* render_backup_selection
Signed-off-by: Aaron Lauterer
---
v1->v2: rebased
www/manager6/Utils.js | 61 +++
www/manager6/dc/Backup
The new detail view for backup jobs shows the settings similar to the
edit dialog but read only. Additionally it does show a list of all
included guests with their volumes and whether these volumes will be
included in the backup.
Signed-off-by: Aaron Lauterer
---
v1->v2:
* m
Adds a new `Summary` button to the backup panel which opens a new
window containing a searchable view of all guests and whether they are
backed up by any job or not.
Signed-off-by: Aaron Lauterer
---
I took the code for the search field from the ResourceGrid, simplified
it and adapted it more
This patch adds a new API endpoint that returns a list of included
guests, their volumes and whether they are included in a backup.
The output is formatted to be used with the extJS tree panel.
Signed-off-by: Aaron Lauterer
---
v1 -> v2:
* simplified the code
* refactored according to feedb
a good one.
This series relies on [1], which is why this is marked as RFC as the one
method call needed from that series will definitely change.
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-January/041268.html
[1] https://pve.proxmox.com/pipermail/pve-devel/2020-April/042753.html
Aaron
On 4/6/20 1:23 PM, Thomas Lamprecht wrote:
On 4/3/20 4:11 PM, Aaron Lauterer wrote:
This extracts the logic which guests are to be included in a backup job
into its own method 'get_included_guests'. This makes it possible to
develop other features around backup jobs.
Logic which was spread
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v3 -> v4: nothing changed. in here for completness sake
src/PVE/
they are to be able to handle missing
permissions according to the current context. The old behavior to die
on a backup job when the user is missing the permission to a guest and
the job is not an 'all' or 'exclude' job is kept.
Signed-off-by: Aaron Lauterer
---
v3 -> v4:
* reworked the whole lo
guest include logic in manager
* fix changed function call in QemuConfig.pm
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-March/042420.html
manager: Aaron Lauterer (1):
backup: move logic to include guests into method
PVE/API2/VZDump.pm | 43 ++
PVE/VZDump.pm
Signed-off-by: Aaron Lauterer
---
v3 -> v4: nothing changed. in here for completness sake
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/src/PVE/LXC/Config.pm b/src/PVE/LXC/Config.pm
index 6cc4e4a..e3be23c 100644
--- a/
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v3->v4: changed function call for `foreach_drive` to QemuServer::Dr
there is no calculation performed and
thus we need to explicitly cast the size to an integer before returning
it.
Signed-off-by: Aaron Lauterer
---
I don't think we need to worry about other ZFS implementations regarding
ZFS over iSCSI. FreeBSD supports it since 9.3 [0], released mid 2014.
Illumos
On 3/27/20 12:22 PM, Thomas Lamprecht wrote:
On 3/16/20 11:02 AM, Aaron Lauterer wrote:
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
Thx for the info :)
On 3/24/20 10:16 AM, Fabian Ebner wrote:
On 17.03.20 15:34, Fabian Grünbichler wrote:
On March 16, 2020 4:44 pm, Aaron Lauterer wrote:
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes
clear
* adding one more item to the FAQ section to troubleshoot
* fix small grammer and sentence structures.
* phrasing some parts in a simpler fashion
Signed-off-by: Aaron Lauterer
---
v1 -> v2:
* remove troubleshooting for failed corosync-qdevice start
this should be fixed with the repacka
Well, as it turns out, the whole (N-1) section is valid and our QDevice
setup code had a bug. Will send a v2 with that section integrated.
On 3/24/20 3:52 PM, Aaron Lauterer wrote:
Naming the whole mechanism and one of the daemons the same makes it easy
to mix up the two. This patch aims
hash it did show
the hash reference. Dereferencing it will again return the number of
items present in the hash.
[0]
https://perldoc.perl.org/perl5260delta.html#scalar(%25hash)-return-signature-changed
Signed-off-by: Aaron Lauterer
---
@Thomas feel free to correct the commit message should my
clear
* removing the `(N-1) qdevice votes` section in odd clusters because it
does not behave like that anymore. Only 1 vote is provided.
* adding two more items to the FAQ section to troubleshoot
* fix small grammer and sentence structures.
Signed-off-by: Aaron Lauterer
---
I had this patch
Looks good to me.
Reviewed-by: Aaron Lauterer
On 3/17/20 1:58 PM, Wolfgang Link wrote:
Improvement of grammar and punctuation.
Clarify the HA limitations.
Remove future tense in some sentences.
It is not good to use it in technical/scientific papers.
Rewrite some sentences to improve
On 3/18/20 2:10 PM, Fabian Grünbichler wrote:
On March 18, 2020 11:57 am, Aaron Lauterer wrote:
On 3/17/20 3:33 PM, Fabian Grünbichler wrote:
On March 16, 2020 4:44 pm, Aaron Lauterer wrote:
This extracts the logic which guests are to be included in a backup job
into its own method
On 3/17/20 3:33 PM, Fabian Grünbichler wrote:
On March 16, 2020 4:44 pm, Aaron Lauterer wrote:
This extracts the logic which guests are to be included in a backup job
into its own method 'get_included_guests'. This makes it possible to
develop other features around backup jobs.
Logic which
new rebased series it out. I did it for each repo to be rather on the
overdone side than missing something :)
On 3/16/20 1:33 PM, Aaron Lauterer wrote:
On 3/16/20 1:16 PM, Fabian Grünbichler wrote:
needs a rebase (please always rebase before sending - the last commit on
src/PVE/VZDump
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v2 -> v3: rebased
v1 -> v2: implemented the suggestion from Fab
Signed-off-by: Aaron Lauterer
---
v2->v3:
rebased
v1->v2:
moved the determination of the reason to the same method that decides if
a mountpoint is included in the backup.
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/s
to the place where it is used.
Signed-off-by: Aaron Lauterer
---
changes: rebased
PVE/API2/VZDump.pm | 36 +++---
PVE/VZDump.pm | 74 --
2 files changed, 55 insertions(+), 55 deletions(-)
diff --git a/PVE/API2/VZDump.pm b/PVE/API2
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v2 -> v3: rebased
v1 -> v2:
* implemented the suggestions from Fab
-February/041972.html
[2] https://pve.proxmox.com/pipermail/pve-devel/2020-March/042258.html
pve-manager: Aaron Lauterer (1):
backup: move logic to include guests into method
PVE/API2/VZDump.pm | 36 +++---
PVE/VZDump.pm | 74 --
2
On 3/16/20 1:16 PM, Fabian Grünbichler wrote:
needs a rebase (please always rebase before sending - the last commit on
src/PVE/VZDump/LXC.pm was done before you sent this patch out!)
Sorry. I'll send a rebased series later today.
___
pve-devel
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
On 3/16/20 8:04 AM, Fabian Grünbichler wrote:
On March 13, 2020 4:21 pm, Aaron Lauterer wrote:
Add the `,audiodev=` property to audio devices when machine version is
4.2 or higher.
With Qemu 4.2 a new `audiodev` property was introduced [0] to explicitly
specify the backend to be used
is
deprecated.
[0] https://wiki.qemu.org/ChangeLog/4.2#Audio
Signed-off-by: Aaron Lauterer
---
tested this with all 3 audio devices available in a local VM and one
live migration from a qemu 4.1 node to a 4.2 node with the default
ich9-intel-hda dev.
PVE/QemuServer.pm | 12 +---
1 file
LGTM thanks :)
On 3/10/20 12:20 PM, Thomas Lamprecht wrote:
On 2/27/20 11:01 AM, Aaron Lauterer wrote:
Signed-off-by: Aaron Lauterer
---
v1[0] -> v2:
given the feedback from Fabian I
* changed name of the method
* added more details as to what implementations should return
[0] ht
Looks good to me. I encountered two things, one missed insert and a
misspelled you->your that I introduced in my comment on the v1. Sorry
for that :/
I marked them inline
Did some changes get dropped or applied in another patch?
The v1 had some changes after line 66. They seem to be missing in
On 3/4/20 3:31 PM, Thomas Lamprecht wrote:
On 2/27/20 12:01 PM, Stefan Reiter wrote:
On 2/27/20 11:48 AM, Fabian Grünbichler wrote:
On February 26, 2020 11:53 am, Aaron Lauterer wrote:
If IO-Thread is activated and a new enough Qemu version installed the
program still ran into the elsif
On 2/27/20 11:01 AM, Aaron Lauterer wrote:
Signed-off-by: Aaron Lauterer
---
v1[0] -> v2:
given the feedback from Fabian I
* changed name of the method
* added more details as to what implementations should return
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-January/041267.h
If active, the return value was a string: "1" and not an integer.
Signed-off-by: Aaron Lauterer
---
PVE/Storage/ISCSIPlugin.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/PVE/Storage/ISCSIPlugin.pm b/PVE/Storage/ISCSIPlugin.pm
index 131ffa0..bb27c89 100644
Signed-off-by: Aaron Lauterer
---
PVE/Storage/ISCSIPlugin.pm | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/PVE/Storage/ISCSIPlugin.pm b/PVE/Storage/ISCSIPlugin.pm
index bb27c89..30b1ceb 100644
--- a/PVE/Storage/ISCSIPlugin.pm
+++ b/PVE/Storage
Move the logic which mountpoints are included in the backup job to its
own method and adapt the VZDump code accordingly. This makes it possible
to develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v1 -> v2: implemented the suggestions from Fabian [0]
[0] ht
in the backup
The behavior of VZDump should still be the same.
[0] https://pve.proxmox.com/pipermail/pve-devel/2020-January/041268.html
pve-manager: Aaron Lauterer (1):
backup: move logic to include guests into method
PVE/API2/VZDump.pm | 36 +++---
PVE/VZDump.pm | 74
Move the logic which volumes are included in the backup job to its own
method and adapt the VZDump code accordingly. This makes it possible to
develop other features around backup jobs.
Signed-off-by: Aaron Lauterer
---
v1 -> v2:
* implemented the suggestions from Fabian [0]
* incorpora
Signed-off-by: Aaron Lauterer
---
moved the determination of the reason to the same method that decides if
a mountpoint is included in the backup.
src/PVE/LXC/Config.pm | 24 +---
1 file changed, 17 insertions(+), 7 deletions(-)
diff --git a/src/PVE/LXC/Config.pm b/src/PVE
1 - 100 of 341 matches
Mail list logo
