Re: [pve-devel] [PATCH ha-manager] always queue service stop if node shuts down

2017-05-26 Thread Dietmar Maurer
Pleas can you resend the whole patch? > This hunk is not needed (but does not affects anything here, as visible > from the regression tests). > Sorry, unclean working directory on commit message re-edit. > ___ pve-devel mailing list pve-devel@pve.pro

Re: [pve-devel] FreeNAS storage plugin

2017-05-26 Thread Michael Rasmussen
The plugin will provide zfs over iscsi. For containers via nfs you can just use the current nfs in freenas. On May 26, 2017 10:53:33 PM GMT+02:00, Andreas Steinel wrote: >Hi mir, > >I'd really love to have ZFS-over-NFS for containers. I'd like to >assist if you like and if you can explain how y

Re: [pve-devel] FreeNAS storage plugin

2017-05-26 Thread Andreas Steinel
Hi mir, I'd really love to have ZFS-over-NFS for containers. I'd like to assist if you like and if you can explain how you setup your test environment. Best, Andreas Steinel (LnxBil) On Fri, May 26, 2017 at 6:20 PM, Michael Rasmussen wrote: > Yes, I intend to upstream this plugin. > > On May 26

Re: [pve-devel] FreeNAS storage plugin

2017-05-26 Thread Michael Rasmussen
Yes, I intend to upstream this plugin. On May 26, 2017 9:09:33 AM GMT+02:00, "Fabian Grünbichler" wrote: >On Wed, May 24, 2017 at 12:58:28PM +0200, datanom.net wrote: >> Hi all, >> >> I have started developing a storage plugin for FreeNAS which uses the >> FreeNAS REST API. >> I have tested thi

[pve-devel] [PATCH ha-manager v2] always queue service stop if node shuts down

2017-05-26 Thread Thomas Lamprecht
Commit 61ae38eb6fc5ab351fb61f2323776819e20538b7 which ensured that services get freezed on a node reboot had a side effect where running services did not get gracefully shutdown on node reboot. This may lead to data loss as the services then get hard killed, or they may even prevent a node reboot b

Re: [pve-devel] [PATCH ha-manager] always queue service stop if node shuts down

2017-05-26 Thread Thomas Lamprecht
On 05/26/2017 04:49 PM, Fabian Grünbichler wrote: small comment inside On Fri, May 26, 2017 at 03:45:38PM +0200, Thomas Lamprecht wrote: [snip] +if ($shutdown && !$reboot) { + $haenv->log('info', "shutdown LRM, stop all services"); + $self->{mode} = 'shutdown'; } else

Re: [pve-devel] [PATCH ha-manager] always queue service stop if node shuts down

2017-05-26 Thread Fabian Grünbichler
small comment inside On Fri, May 26, 2017 at 03:45:38PM +0200, Thomas Lamprecht wrote: > Commit 61ae38eb6fc5ab351fb61f2323776819e20538b7 which ensured that > services get freezed on a node reboot had a side effect where running > services did not get gracefully shutdown on node reboot. > This may

Re: [pve-devel] [PATCH pve-manager 05/18] PVE::Replication - use new calendar events instead of interval

2017-05-26 Thread Fabian Grünbichler
comment inline On Tue, May 23, 2017 at 09:08:44AM +0200, Dietmar Maurer wrote: > And implement retry algorythm after failure: > > $next_sync = $state->{last_try} + 5*60*$fail_count; > > and limit to 3 failures. > > Signed-off-by: Dietmar Maurer > --- > PVE/API2/Replication.pm | 2 +- > PVE

Re: [pve-devel] [PATCH ha-manager] always queue service stop if node shuts down

2017-05-26 Thread Thomas Lamprecht
On 05/26/2017 03:45 PM, Thomas Lamprecht wrote: [..snip..] diff --git a/src/PVE/HA/NodeStatus.pm b/src/PVE/HA/NodeStatus.pm index b99689c..493298f 100644 --- a/src/PVE/HA/NodeStatus.pm +++ b/src/PVE/HA/NodeStatus.pm @@ -221,8 +221,12 @@ sub fence_node { if ($success) { my $msg = "f

Re: [pve-devel] [PATCH pve-manager 01/18] pvesr: add pve storage replication tool

2017-05-26 Thread Fabian Grünbichler
comments inline On Tue, May 23, 2017 at 09:08:40AM +0200, Dietmar Maurer wrote: > Just added code to configure jobs. Replication itself is not > implemented. > > Signed-off-by: Dietmar Maurer > --- > PVE/API2/Cluster.pm | 7 ++ > PVE/API2/Makefile | 2 + > PVE/API2/Nod

Re: [pve-devel] [PATCH pve-guest-common 1/2] PVE/ReplicationConfig.pm - new configuratin file for replication

2017-05-26 Thread Fabian Grünbichler
meta-remark: IMHO this should be 2 or even 3 files, not one. at least the base config and the cluster config should be split. more remarks inline On Tue, May 23, 2017 at 08:40:48AM +0200, Dietmar Maurer wrote: > Signed-off-by: Dietmar Maurer > --- > Makefile | 1 + > PVE/Repl

[pve-devel] [PATCH ha-manager] always queue service stop if node shuts down

2017-05-26 Thread Thomas Lamprecht
Commit 61ae38eb6fc5ab351fb61f2323776819e20538b7 which ensured that services get freezed on a node reboot had a side effect where running services did not get gracefully shutdown on node reboot. This may lead to data loss as the services then get hard killed, or they may even prevent a node reboot b

[pve-devel] [PATCH spiceterm] fix memory leaks when using g_hashtable

2017-05-26 Thread Dominik Csapak
when generating a bitmap for a character whose codepoint was above 255, we used a cache_id of 0, malloc'd a CachedImage and inserted it into a g_hashtable, without freeing the one which was before inserted with cache_id 0 this is circumvented by only generating a CachedImage when having a cache_id

Re: [pve-devel] [PVE-User] Proxmox VE 5.0 beta2 released!

2017-05-26 Thread Gilberto Nunes
Hi Alexandre I follow your instruction, but when I try to start the VM ( after set VM to boot with CDROM add but cloudinit), and I get the error: May 26 07:39:27 qapla pvedaemon[5821]: unable to find free nbd device This is the VM conf: cat /etc/pve/qemu-server/113.conf boot: dcn bootdisk: virt

Re: [pve-devel] FreeNAS storage plugin

2017-05-26 Thread Fabian Grünbichler
On Wed, May 24, 2017 at 12:58:28PM +0200, datanom.net wrote: > Hi all, > > I have started developing a storage plugin for FreeNAS which uses the > FreeNAS REST API. > I have tested this using a combination of Postman, iscsi_discovery, and > iscsi-inq and everything seems to work perfectly. Using t