[pve-devel] [PATCH] prepare code for more generic firewall logging

2017-09-14 Thread Tom Weber
making ruleset generation aware of a match and action part in iptable rules. code will generate the same iptables as before! (except for a few additional spaces between match and action). ---  src/PVE/Firewall.pm | 168 +++-  1 file changed, 99 inserti

[pve-devel] [PATCH] Firewall Improvements

2017-09-14 Thread Tom Weber
See mail Firewall Improvements Tom Weber (1):   prepare code for more generic firewall logging  src/PVE/Firewall.pm | 168 +++-  1 file changed, 99 insertions(+), 69 deletions(-) ___ pve-devel mailing lis

[pve-devel] Firewall Improvements

2017-09-14 Thread Tom Weber
Hi all, last week I reported a problem with firewall logging. After looking deeper into Firewall.pm I have a better understanding of the problems I first had with using the Firewall as a rather fresh PVE User: - the different levels of log_level_in / out don't make sense to me. Firewall.pm uses t

[pve-devel] applied: [PATCH qemu-server 1/1] do not overwrite global signal handlers

2017-09-14 Thread Wolfgang Bumiller
applied On Thu, Sep 14, 2017 at 03:19:39PM +0200, Emmanuel Kasper wrote: > perls 'local' must be either used in front of each $SIG{...} > assignments or they must be put in a list, else it affects only the > first variable and the rest are *not* in local context. > > In all cases the global signa

[pve-devel] [PATCH qemu-server 1/1] do not overwrite global signal handlers

2017-09-14 Thread Emmanuel Kasper
perls 'local' must be either used in front of each $SIG{...} assignments or they must be put in a list, else it affects only the first variable and the rest are *not* in local context. In all cases the global signal handlers we overwrote were in cli programs or forked workers, not in daemons. ---

[pve-devel] [PATCH qemu-server 0/1] Do not overwrite global signal handlers, extra fix

2017-09-14 Thread Emmanuel Kasper
While poking around QemuServer.pm I noticed a couple of places where we were using 'local' in a wrong way like Thomas reported. After checking with grep in /usr/share/perl5/PVE, this should be now fixed everywhere. Emmanuel Kasper (1): do not overwrite global signal handlers PVE/API2/Qemu.pm

[pve-devel] [RFC access-control] API/ticket: rework coarse grained permission computation

2017-09-14 Thread Thomas Lamprecht
We accessed methods from PVE::Storage here but did not define a "use PVE::Storage". This thus only worked if modules if the PVE::Storage module got pulled in by something else, by luck. Simply including said use statement is not an option because pve-storage is already dependent from pve-access-con

[pve-devel] [PATCH ha-manager v2] fix #1347: let postfix fill in FQDN in fence mails

2017-09-14 Thread Philip Abernethy
Using the nodename in $mailto is not correct and can lead to mails not forwarding in restrictive mail server configurations. Also changes $mailfrom to 'root' instead of 'root@localhost', which results in postfix appending the proper FQDN there, too. As a result the Delivered-to header reads somethi

[pve-devel] [PATCH manager v2 1/2] vzdump: handle backup to stdout correctly

2017-09-14 Thread Philip Abernethy
Checks for dumpdir or tmpdir backups don't apply to stdout, so we can skip the checks involved in those methods. --- changes from v1: * Added an explanation for tmpdir. I'm not sure there's much more we can do. If the user/admin configures the backup to save a 30G container into 16G memory he's

[pve-devel] [PATCH manager v2 2/2] vzdump: whitespace and style cleanup, typo fix

2017-09-14 Thread Philip Abernethy
Mostly whitespace fixes, a stray semicolon, some style fixes and a grammar correction --- PVE/VZDump.pm | 111 -- 1 file changed, 53 insertions(+), 58 deletions(-) diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm index c6a4b449..844a9449 100644 --

Re: [pve-devel] [PATCH ha-manager] fix #1347: let postfix fill in FQDN in fence mails

2017-09-14 Thread Thomas Lamprecht
On 09/14/2017 11:15 AM, Philip Abernethy wrote: On Wed, 2017-09-13 at 08:11 +0200, Thomas Lamprecht wrote: On 09/08/2017 02:56 PM, Philip Abernethy wrote: Using the nodename is not correct and can lead to mails not forwarding in restrictive mail server configurations. ---   src/PVE/HA/Env/PVE2

Re: [pve-devel] [PATCH ha-manager] fix #1347: let postfix fill in FQDN in fence mails

2017-09-14 Thread Philip Abernethy
On Wed, 2017-09-13 at 08:11 +0200, Thomas Lamprecht wrote: > On 09/08/2017 02:56 PM, Philip Abernethy wrote: > > Using the nodename is not correct and can lead to mails not > > forwarding > > in restrictive mail server configurations. > > --- > >   src/PVE/HA/Env/PVE2.pm | 4 ++-- > >   1 file chang

Re: [pve-devel] [PATCH cluster v2 5/7] corosync: add atomic_write_conf and cleanup

2017-09-14 Thread Thomas Lamprecht
On 09/14/2017 10:18 AM, Fabian Grünbichler wrote: On Thu, Sep 14, 2017 at 10:00:39AM +0200, Thomas Lamprecht wrote: Signed-off-by: Thomas Lamprecht --- data/PVE/Corosync.pm | 39 --- 1 file changed, 16 insertions(+), 23 deletions(-) diff --git a/data/PVE/

Re: [pve-devel] [PATCH cluster v2 5/7] corosync: add atomic_write_conf and cleanup

2017-09-14 Thread Fabian Grünbichler
On Thu, Sep 14, 2017 at 10:00:39AM +0200, Thomas Lamprecht wrote: > Signed-off-by: Thomas Lamprecht > --- > data/PVE/Corosync.pm | 39 --- > 1 file changed, 16 insertions(+), 23 deletions(-) > > diff --git a/data/PVE/Corosync.pm b/data/PVE/Corosync.pm > index

Re: [pve-devel] RFC bump Version to 2.8.0 OpenvSwitch

2017-09-14 Thread Alexandre DERUMIER
Seem they have bumped dpdk version in ovs 2.8 :( http://docs.openvswitch.org/en/latest/faq/releases/ Open vSwitchDPDK 2.2.x 1.6 2.3.x 1.6 2.4.x 2.0 2.5.x 2.2 2.6.x 16.07.2 2.7.x 16.11.2 2.8.x 17.05.1 17.05 Available in ubuntu artful. https://packages.ubuntu.com/artful/dpdk T

[pve-devel] [PATCH cluster v2 2/7] pvecm: import often needed run_command

2017-09-14 Thread Thomas Lamprecht
Signed-off-by: Thomas Lamprecht --- data/PVE/CLI/pvecm.pm | 16 +++- 1 file changed, 7 insertions(+), 9 deletions(-) diff --git a/data/PVE/CLI/pvecm.pm b/data/PVE/CLI/pvecm.pm index 41c2f9c..5a93a66 100755 --- a/data/PVE/CLI/pvecm.pm +++ b/data/PVE/CLI/pvecm.pm @@ -10,7 +10,7 @@ use

[pve-devel] [PATCH cluster v2 4/7] corosync: transform config to allow easier access

2017-09-14 Thread Thomas Lamprecht
Signed-off-by: Thomas Lamprecht --- data/PVE/Corosync.pm | 64 +--- 1 file changed, 36 insertions(+), 28 deletions(-) diff --git a/data/PVE/Corosync.pm b/data/PVE/Corosync.pm index 3c4c8c0..1180316 100644 --- a/data/PVE/Corosync.pm +++ b/data/PVE/C

[pve-devel] [PATCH cluster v2 7/7] WIP: add and delete qdevice

2017-09-14 Thread Thomas Lamprecht
Allow to add and delete qdevice through the pvecm CLI tool. Signed-off-by: Thomas Lamprecht --- NOTE: WIP data/PVE/CLI/pvecm.pm | 208 ++ 1 file changed, 208 insertions(+) diff --git a/data/PVE/CLI/pvecm.pm b/data/PVE/CLI/pvecm.pm index 5a93a66.

[pve-devel] [PATCH cluster v2 1/7] pvecm: remove Data::Dumper

2017-09-14 Thread Thomas Lamprecht
Signed-off-by: Thomas Lamprecht --- data/PVE/CLI/pvecm.pm | 1 - 1 file changed, 1 deletion(-) diff --git a/data/PVE/CLI/pvecm.pm b/data/PVE/CLI/pvecm.pm index 7ca98f7..41c2f9c 100755 --- a/data/PVE/CLI/pvecm.pm +++ b/data/PVE/CLI/pvecm.pm @@ -10,7 +10,6 @@ use POSIX; use Net::IP; use File::Pa

[pve-devel] [PATCH cluster v2 3/7] corosync config parser: move to hash format

2017-09-14 Thread Thomas Lamprecht
The old parser itself was simple and easy but resulted in quite a bit of headache when changing corosync config sections, especially if multiple section levelsshould be touched. Move to a more practical internal format which represents the corosync configuration in hash Signed-off-by: Thomas Lamp

[pve-devel] [PATCH cluster v2 0/7][RESEND] corosync config parser refactoring and a WIP of QDevice

2017-09-14 Thread Thomas Lamprecht
This is a resend of my last iteration, code wise nothing changed but it did not apply on current master, so I respinned it. First 5 patches should be ready for comitting, but the last two are rather RFC/WIP and for people who'd like to test (easy) qdevice setup now. But if someone gives them a (hi

[pve-devel] [PATCH cluster v2 6/7] WIP: add foreach_member

2017-09-14 Thread Thomas Lamprecht
Add a method to execute a sub for each cluster member, passing its name and cluster ip to the $code sub. This is similar to our foreach_disk or foreach_storage Signed-off-by: Thomas Lamprecht --- NOTE: WIP, and only for the next patchs qdevice prototype data/PVE/Cluster.pm | 16 +++

[pve-devel] [PATCH cluster v2 5/7] corosync: add atomic_write_conf and cleanup

2017-09-14 Thread Thomas Lamprecht
Signed-off-by: Thomas Lamprecht --- data/PVE/Corosync.pm | 39 --- 1 file changed, 16 insertions(+), 23 deletions(-) diff --git a/data/PVE/Corosync.pm b/data/PVE/Corosync.pm index 1180316..1d58bf0 100644 --- a/data/PVE/Corosync.pm +++ b/data/PVE/Corosync.pm @@

Re: [pve-devel] RFC bump Version to 2.8.0 OpenvSwitch

2017-09-14 Thread Wolfgang Link
> Alexandre DERUMIER hat am 14. September 2017 um 09:16 > geschrieben: > > > Hi, > > could it be possible to add dpdk support ? I did not try it but the manual has a requirement for DPDK Version17.05.1 http://docs.openvswitch.org/en/latest/intro/install/dpdk/ Debian has only 16.11.1 in stre

Re: [pve-devel] RFC bump Version to 2.8.0 OpenvSwitch

2017-09-14 Thread Alexandre DERUMIER
Hi, could it be possible to add dpdk support ? I had sent a patch some months ago https://pve.proxmox.com/pipermail/pve-devel/2017-April/026089.html It's only add support to dpdk, but don't enable it by default, so I think it's quite safe. - Mail original - De: "Wolfgang Link" À: "p

Re: [pve-devel] [PATCH qemu-server v8 1/6] Force overwriting existing compressed man pages and symbolic links

2017-09-14 Thread Emmanuel Kasper
On 09/14/2017 08:09 AM, Thomas Lamprecht wrote: > On 09/13/2017 04:10 PM, Emmanuel Kasper wrote: >> This allows calling the 'make install' target twice in a row. > > Looks OK, we overwrite everything else so why not the man pages. > > This can be already applied even if we would say the current

[pve-devel] applied: [PATCH installer] fix #1479: detect missing KVM support

2017-09-14 Thread Wolfgang Bumiller
applied On Tue, Aug 22, 2017 at 02:18:46PM +0200, Fabian Grünbichler wrote: > Signed-off-by: Fabian Grünbichler > --- > proxinstall | 8 > 1 file changed, 8 insertions(+) > > diff --git a/proxinstall b/proxinstall > index 509430f..a04d98a 100755 > --- a/proxinstall > +++ b/proxinstall

[pve-devel] applied: [RFC cluster] update SSH Ciphers for Debian Stretch

2017-09-14 Thread Wolfgang Bumiller
applied On Wed, May 31, 2017 at 09:38:00AM +0200, Fabian Grünbichler wrote: > blowfish, 3des and arcfour are not enabled by default on the > server side anyway. > > on most hardware, AES is about 3 times faster than Chacha20 > because of hardware accelerated AES, hence the changed order > of pref