Re: [PVE-User] ZFS on iSCSI + Pacemaker/corosync/DRBD

2016-11-23 Thread Michael Rasmussen
On Wed, 23 Nov 2016 09:40:55 + Mark Adams wrote: > > Has anyone else tried to get this or a similar setup working? Any views > greatly received. > What you are trying to achieve is not a good idea with corosync/pacemaker since iSCSI is a block device. To create a cluster over a LUN will req

Re: [PVE-User] MTU size changed on a running cluster

2016-11-23 Thread Wolfgang Link
This is a openvswitch bug. The workaround is to use openvswitch 2.6, it is on testing repo and set mtu_reqest on the interface. https://github.com/openvswitch/ovs/blob/master/FAQ.rst Q: How can I configure the bridge internal interface MTU? Why does Open vSwitch keep changing internal ports MTU?

[PVE-User] MTU size changed on a running cluster

2016-11-23 Thread Marco M. Gabriel
Hi there, on a productive 5 node Proxmox VE Ceph cluster, we experienced some strange behaviour: Based on http://pve.proxmox.com/wiki/Open_vSwitch#Example_2:_Bond_.2B_Bridge_.2B_Internal_Ports we have an internal network for cluster/corosync communication and another internal network for Ceph Sto

Re: [PVE-User] A stop job is running... (xxx/no limit)

2016-11-23 Thread Alain Péan
Le 22/11/2016 à 19:00, Gerald Brandt a écrit : I'm trying to shut down a server, and it waits on 'A stop job is running... (xx/ no limit). Why is there no time limit, and how can I set one? I see also this problem with a Dell R630 server with Broadcom 10g interfaces. It's a known bug. It is r

Re: [PVE-User] A stop job is running... (xxx/no limit)

2016-11-23 Thread Marco Gaiarin
Mandi! Gerald Brandt In chel di` si favelave... > >NFS storage? > Yup. Why, does that make a difference? Look at list archive, some weeks ago: seems that systemd behave not so correctly and tear down the NFS server before proxmox, that stalls. -- dott. Marco Gaiarin

Re: [PVE-User] A stop job is running... (xxx/no limit)

2016-11-23 Thread Gerald Brandt
On 2016-11-23 06:30 AM, Marco Gaiarin wrote: Mandi! Gerald Brandt In chel di` si favelave... I'm trying to shut down a server, and it waits on 'A stop job is running... (xx/ no limit). Why is there no time limit, and how can I set one? NFS storage? Yup. Why, does that make a difference

Re: [PVE-User] A stop job is running... (xxx/no limit)

2016-11-23 Thread Marco Gaiarin
Mandi! Gerald Brandt In chel di` si favelave... > I'm trying to shut down a server, and it waits on 'A stop job is > running... (xx/ no limit). Why is there no time limit, and how can I > set one? NFS storage? -- dott. Marco Gaiarin GNUPG Key ID: 240A3D66

[PVE-User] ZFS on iSCSI + Pacemaker/corosync/DRBD

2016-11-23 Thread Mark Adams
Hi All, I'm testing out proxmox and trying to get a working ZFS on iSCSI HA setup going. Because ZFS on iSCSI logs on to the iscsi server via ssh and creates a zfs dataset then adds iscsi config to /etc/ietd.conf it works fine when you've got a single iscsi host, but I haven't figured out a way t