>>And I still wonder it it would be possible to add NF_HOOKS
>>to tun and veth devices directly? Or is that a bad idea?
I really don't known. (How does it work ?).
- Mail original -
De: "dietmar"
À: "aderumier" , "pve-devel"
Envoyé: Vendredi 4 Mars 2016 07:45:43
Objet: Re: [pve-devel]
> > ovs 2.5 support conntrack and openstack have commited their ovs firewall (no
> > iptables, no bridge on top of ovs)
>
> Oh, that is based on openflow - means we would need to
> have 2 totally different firewall implementations?
And I still wonder it it would be possible to add NF_HOOKS
to tu
>>Oh, that is based on openflow - means we would need to
>>have 2 totally different firewall implementations?
Yes. only kernel conntrack is common.
For now, I think it's not a priority.That need a lot of work, and
I'm not sure all features (ipset) are available.
(I don't known too much how open
> ovs 2.5 support conntrack and openstack have commited their ovs firewall (no
> iptables, no bridge on top of ovs)
Oh, that is based on openflow - means we would need to
have 2 totally different firewall implementations?
___
pve-devel mailing list
pve
> ovs 2.5 support conntrack and openstack have commited their ovs firewall (no
> iptables, no bridge on top of ovs)
Yes, that would be an interesting optimization. I guess we can get rid of the
additional bridge?
If so, can we do the same for linux bridge?
__
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied all 3 patches
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
> > There should be a way for files to be discarded but the dir structure to be
> > kept. I know tar has the --exclude-tag-under for example, and google gives
> > funky joined tar and "find -type d" invocation to deal with this... any
> > thoughts?
>
> Definitely sounds like a good idea. The hard
Also,
ovs 2.5 support conntrack and openstack have commited their ovs firewall (no
iptables, no bridge on top of ovs)
https://git.openstack.org/cgit/openstack/neutron/commit/?id=ef29f7eb9a2a37133eacdb7f019b48ec3f9a42c3
- Mail original -
De: "aderumier"
À: "pve-devel"
Envoyé: Jeudi 3
http://openvswitch.org/releases/NEWS-2.5.0
I'll try to test it soon, I was waiting for it to test DPDK and vhost-user
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
> On March 3, 2016 at 3:58 PM Jean-Francois Dagenais
> wrote:
>
>
> While we are on the subject of vzdump. One thing that fails when restoring
> dumps
> is the lack of the original directory structure. Here's an example:
>
> /var/log/?* is excluded, however, there are subdirectories in there
Looks good. Now since these patches today seem to be your first
contributions to PVE there's one more thing left to do, and
that's the license agreement (AGPLv3 and the Harmony CLA), see
the bottom of the developer documentation in our wiki:
http://pve.proxmox.com/wiki/Developer_Documentation
We
While we are on the subject of vzdump. One thing that fails when restoring dumps
is the lack of the original directory structure. Here's an example:
/var/log/?* is excluded, however, there are subdirectories in there which were
carefully created by a package managers (APT or RPM) at package instal
Not all document default values matched their actual
default, fix this by not maintaining two separate lists of
default values.
Note that this changes:
- the default compression from none to the documented LZO
- the documented default mode from stop to snapshot
---
It's a compromise, I felt LZO
> On March 3, 2016 at 3:05 PM Jean-Francois Dagenais
> wrote:
>
>
>
> > On Mar 3, 2016, at 03:53, Wolfgang Bumiller wrote:
> >
> >>> Also, the stdexcludes is broken as well. The block starting with "if
> >>> ($opts->{stdexcludes}) {" is not executed even though it should based on
> >>> the
Previously, foreach_drive iterated over all configuration
keys (in a random order) and checked whether the current key
is a valid drive name. Instead, we now iterate over a list
of valid drive names (with deterministic order) and check
whether a drive with such a name exists in the
configuration.
and not files like "acpid"
Signed-off-by: Jean-Francois Dagenais
---
PVE/VZDump.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm
index 8f5af3f..a3d1f2c 100644
--- a/PVE/VZDump.pm
+++ b/PVE/VZDump.pm
@@ -593,7 +593,7 @@ sub new {
push @$
> On Mar 3, 2016, at 03:22, Wolfgang Bumiller wrote:
>
> Call split_args() on the 'exclude-path' option read from
> /etc/vzdump.conf
>
> With the move to the generic JSONSchema::parse_config() in
> commit 16145cba the call to split_args() for 'exclude-path'
> was dropped breaking the 'exclude-p
> On March 3, 2016 at 3:05 PM Wolfgang Bumiller wrote:
>
>
> > On March 3, 2016 at 2:48 PM Jean-Francois Dagenais
> > wrote:
> > > On Mar 3, 2016, at 03:55, Wolfgang Bumiller
> > > wrote:
> (...)
> > > + '/var/run/?*pid';
> >
> > Shouldn't this be "/var/run/?*.pid" ? I'm th
> On March 3, 2016 at 2:48 PM Jean-Francois Dagenais
> wrote:
> > On Mar 3, 2016, at 03:55, Wolfgang Bumiller wrote:
(...)
> > +'/var/run/?*pid';
>
> Shouldn't this be "/var/run/?*.pid" ? I'm thinking of file names ending
> in "pid" such as "acpid" for example.
No argument
> On Mar 3, 2016, at 03:53, Wolfgang Bumiller wrote:
>
>>> Also, the stdexcludes is broken as well. The block starting with "if
>>> ($opts->{stdexcludes}) {" is not executed even though it should based on the
>>> fact that I don't mention it and it should therefore default to true.
>
> Not sure
> On Mar 3, 2016, at 03:55, Wolfgang Bumiller wrote:
>
> It used to be /var/run/.+pid when we used regular
> expressions so now it should be /var/run/?*pid
> ---
> PVE/VZDump.pm | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm
> index 218b033
our parent class Ext.grid.GridPanel sets 'tbar: null'
and Ext.applyIf() will only apply our property
if it is undefined, so switch to Ext.apply()
it is safe to use Ext.apply() because none of these classes is further extended,
so we don't have to worry that might me might override a toolbar
set fu
---
www/manager6/lxc/Config.js | 110 ++---
1 file changed, 55 insertions(+), 55 deletions(-)
diff --git a/www/manager6/lxc/Config.js b/www/manager6/lxc/Config.js
index a05fb08..8902c86 100644
--- a/www/manager6/lxc/Config.js
+++ b/www/manager6/lxc/Config.j
we used a custom 'data' property to pass around the items to
add to the store, but this property is now used by ExtJS to
store content for the component template
also move to declarative style, saves lines
---
www/manager6/form/Boolean.js| 17 +++-
www/manager6/form/BusTypeSel
we had the same actions for different if paths,
now they are combined
Signed-off-by: Dominik Csapak
---
www/manager/qemu/HDMove.js | 16 +++-
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/www/manager/qemu/HDMove.js b/www/manager/qemu/HDMove.js
index 824a898..6144060 10
if one selected an item in a ComboGrid which was already selected,
the ComboGrid would not close (as expected), this patch fixes this
Signed-off-by: Dominik Csapak
---
www/manager6/form/ComboGrid.js | 12
1 file changed, 12 insertions(+)
diff --git a/www/manager6/form/ComboGrid.js
> Emmanuel Kasper hat am 3. März 2016 um 13:46
> geschrieben:
>
> diff --git a/PVE/API2/Qemu.pm b/PVE/API2/Qemu.pm
> index efac2c7..1bff2f7 100644
> --- a/PVE/API2/Qemu.pm
> +++ b/PVE/API2/Qemu.pm
> @@ -1623,7 +1623,8 @@ __PACKAGE__->register_method({
> method => 'POST',
> protected =>
---
src/PVE/API2/LXC/Status.pm | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/src/PVE/API2/LXC/Status.pm b/src/PVE/API2/LXC/Status.pm
index c5fca29..01dfe12 100644
--- a/src/PVE/API2/LXC/Status.pm
+++ b/src/PVE/API2/LXC/Status.pm
@@ -198,7 +198,7 @@ __PACKAGE__->register_
---
PVE/API2/Qemu.pm | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/PVE/API2/Qemu.pm b/PVE/API2/Qemu.pm
index efac2c7..1bff2f7 100644
--- a/PVE/API2/Qemu.pm
+++ b/PVE/API2/Qemu.pm
@@ -1623,7 +1623,8 @@ __PACKAGE__->register_method({
method => 'POST',
protected
This is necessary, if the volume group "pve" exists, say from a previous
installation.
But whitout printing the reason no user will understand why this happens.
---
proxinstall | 8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/proxinstall b/proxinstall
index ec15477..927
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied all patches - great!
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
call cfs_config_path and parse_lxc_network in
PVE::LXC::Config instead of PVE::LXC
---
Rebased on current master
src/PVE/Firewall.pm | 6 +++---
src/PVE/FirewallSimulator.pm | 2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/src/PVE/Firewall.pm b/src/PVE/Firewall.pm
i
---
PVE/CLI/vzdump.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/PVE/CLI/vzdump.pm b/PVE/CLI/vzdump.pm
index 4684d89..a048a38 100755
--- a/PVE/CLI/vzdump.pm
+++ b/PVE/CLI/vzdump.pm
@@ -126,7 +126,7 @@ vzdump skips the following files wit option --stdexcludes
You can man
It used to be /var/run/.+pid when we used regular
expressions so now it should be /var/run/?*pid
---
PVE/VZDump.pm | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm
index 218b033..8f5af3f 100644
--- a/PVE/VZDump.pm
+++ b/PVE/VZDump.pm
@@ -593,7 +593,
---
PVE/CLI/vzdump.pm | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/PVE/CLI/vzdump.pm b/PVE/CLI/vzdump.pm
index e7dea56..4684d89 100755
--- a/PVE/CLI/vzdump.pm
+++ b/PVE/CLI/vzdump.pm
@@ -119,10 +119,10 @@ You can specify a hook script with option C<--script>.
This s
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
applied both patches
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
> > This made it work again, except I had to do it to defaults and opt->... to
> > get
> > rid of the error. The result is that the excludes are added twice on the
> > tar cmd
> > line.
So this was due to the copying of the defaults read from /etc/vzdump.conf
to $opts when no such option was spe
applied
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
The exclude list is built from both the command line
parameters and the config. If no excludes are passed on the
command line we previously copied the defaults to it,
effectively passing them twice.
---
PVE/VZDump.pm | 1 +
1 file changed, 1 insertion(+)
diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm
applied all 4 patches, thanks!
___
pve-devel mailing list
pve-devel@pve.proxmox.com
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Thanks for reporting, apparently /etc/vzdump.conf has been left
out in the cold for a while...
I just posted a fix for this using PVE::Tools::split_args() (like
it used to be before July) which uses Text::ParseWords::shellwords()
and basically separates the paths the way a shell would do, by
whites
The schema's documentation was already updated to reflect
the change from regular expressions to shell globs, but the
pod entry in CLI/vzdump.pm wasn't.
Also C<> is not interpreted in verbatim (space-prefixed)
lines so it showed up as literal text.
---
PVE/CLI/vzdump.pm | 2 +-
1 file changed, 1
Call split_args() on the 'exclude-path' option read from
/etc/vzdump.conf
With the move to the generic JSONSchema::parse_config() in
commit 16145cba the call to split_args() for 'exclude-path'
was dropped breaking the 'exclude-path' option.
---
PVE/VZDump.pm | 3 +++
1 file changed, 3 insertions(
51 matches
Mail list logo