[ovirt-users] oVirt 3.5 Neutron (Will this work?)

2014-10-25 Thread Assaf Muller
Hello Phil,

The current setup doesn't seem to leverage oVirt's and Neutron's features.
I would try to move the tunnel between the hosts, or some higher-up entity
in your data centers. Would that be possible while adhering to your security 
requirements?
You could then use Neutron FWaaS to replace the VM firewall, and Neutron LBaaS 
to replace OSPF.

 - Forwarded Message -
 From: Phil Daws ux...@splatnix.net
 To: Moti Asayag masa...@redhat.com
 Cc: users users@ovirt.org
 Sent: Tuesday, October 21, 2014 10:21:18 PM
 Subject: Re: [ovirt-users] oVirt 3.5  Neutron (Will this work?)
 
 Hi Moti:
 
 Have thrown together a diagram of how I think it should look :-
 
 https://cloudvault.innoffice247.com/public.php?service=filest=9e57686453ce6b71fdce1dd1eb18fe4a
 
 As each oVirt host has a single activated NIC am trying to see how I can use
 OVS, to provide OSPF  SFLOW, and vLAN capability like I have used manually
 with KVM and OVS. From my dev machine this is how OVS looks:
 
 [root@dev01 ~]# ovs-vsctl show
 55a2af2f-daf5-4f01-a757-9bccaf4f6932
 Bridge ovsbr0
 Port vnet0
 Interface vnet0
 Port vnet1
 tag: 8
 Interface vnet1
 Port vnet13
 tag: 14
 Interface vnet13
 Port vnet9
 tag: 10
 Interface vnet9
 Port mgmt0
 Interface mgmt0
 type: internal
 Port vnet14
 tag: 8
 Interface vnet14
 Port ovsbr0
 Interface ovsbr0
 type: internal
 Port vnet11
 tag: 8
 Interface vnet11
 Port vnet10
 tag: 13
 Interface vnet10
 Port vnet12
 tag: 13
 Interface vnet12
 Port em1
 Interface em1
 Port vnet3
 tag: 14
 Interface vnet3
 Port vnet4
 tag: 20
 Interface vnet4
 Port vnet2
 tag: 10
 Interface vnet2
 ovs_version: 2.3.90
 
 So I have a single NIC with a public facing IP and then I present that IP as
 a gateway, via the bridge, to a VM firewall which then handles the vlans
 inside that.
 
 Hope that makes sense ?
 
 Thanks, Phil
 
 
 - Original Message -
 From: Phil Daws ux...@splatnix.net
 To: Moti Asayag masa...@redhat.com
 Cc: users users@ovirt.org
 Sent: Tuesday, 21 October, 2014 5:26:33 PM
 Subject: Re: [ovirt-users] oVirt 3.5  Neutron (Will this work?)
 
 Hi Moti:
 
 thank you for detailed response. I will diagram what I am thinking as that
 should explain it a whole lot better :)
 
 Thanks, Phil
 
 - Original Message -
 From: Moti Asayag masa...@redhat.com
 To: Phil Daws ux...@splatnix.net
 Cc: users users@ovirt.org
 Sent: Tuesday, 21 October, 2014 4:50:45 PM
 Subject: Re: [ovirt-users] oVirt 3.5  Neutron (Will this work?)
 
 
 Hi Phil,
 
 See answers/questions inline.
 
 - Original Message -
  From: Phil Daws ux...@splatnix.net
  To: users users@ovirt.org
  Sent: Tuesday, October 21, 2014 6:05:55 PM
  Subject: Re: [ovirt-users] oVirt 3.5  Neutron (Will this work?)
  
  Hmmm, this is interesting as it would appear you can only use the Neutron
  appliance with a brand new host ?!? so how does one switch to use it on a
  current system ?
 
 A new host is not mandatory. You need to move an existing host to maintenance
 and reinstall it. On the Re-install dialog select the details of the
 network
 provider.
 
  
  Thanks, Phil
  
  - Original Message -
  From: Phil Daws ux...@splatnix.net
  To: users@ovirt.org
  Sent: Tuesday, 21 October, 2014 1:31:09 PM
  Subject: [ovirt-users] oVirt 3.5  Neutron (Will this work?)
  
  Hello:
  
  have installed oVirt 3.5 on two cloud based servers and then managing them
  from a local engine using a VPN link. On each server I would like to use
  the Neutron VM appliance so that I can provision the networks using
  OpenVswitch as I would like to learn about using OSPF between the two
  diverse systems.
  
 
 Do you intend to use a single neutron appliance for each host or to use a
 single
 neutron appliance to manage connectivity on the two hosts ?
 
  The question is that only physical NIC is enabled (public facing), and
  occupies the ovirtmgmt network, so would I still be able to use Neutron on
  the second NIC even though it is not connected to anything ?
 
 Is there L2 connectivity between the hosts ? Or by not connected to
 anything you
 actually mean there is no wiring between the hosts ?
 
 If this is the first case, you should be able to define for each subnet a
 gateway
 via the 'Add subnet' dialog on the engine. That gateway should be used for
 obtaining
 connectivity for the vms to the public/external network. You'll have to
 configure it manually
 though (doesn't covered as part of the ovirt-neutron integration).
 
 If there is really no connectivity between the hosts and the only outgoing
 traffic from
 each host is via the ovirtmgmt network - it is problematic. According to [1],
 you'll
 have to specify as bridge mappings on the network provider agent details:
 br-neutron:ovirtmgmt,
 where ovirtmgmt will replace the neutron.
 
 But that also implies that you'll share any traffic going through the
 integration bridge
 of neutron and the hosts with the management network and respectively with
 the public
 network, hence the dhcp agents 

Re: [Users] oVirt in PAT - NAT environment

2014-03-10 Thread Assaf Muller
- Original Message -
 Hello folks,
 

Hi!

 I know, this is a topic where you have spend a lots of words, but, after
 installing oVirt 3.3 and 3.4 on CentOS and let goign google search engine to
 become hot, I don´t found any solution for my environment.
 
 My environment:
 
 - dedicated server at server4you.net with CentOS 6
 - Main IP (public) 82.25.xx.yy
 - Addon IP (public) 62.75.xx.yy
 
 What I will do:
 
 - create VMs on a private network 192.x.y.z
 - talk to the internet from the VMs (NAT)
 - reach the VMs from the internet (PAT)
 - reach the VMs via SSH from the host server
 

oVirt doesn't have built-in support for NAT. What you can do is either:
1) Define PAT on the data center edge router. For accessing VMS from the outside
you can either define port forwarding or DNAT. As for reaching the VMs from your
dedicated server, the easiest approach would be ensuring that the server has
a leg in the private network.

2) You might want to try the oVirt Neutron integration. In Neutron you can
create arbitrary private networks and VMs will get IPs from a DHCP server
created in Neutron. You can then define a 'floating ip', which is a mapping
between the VM's private IP and an allocated public IP on your public network.
Neutron will do the DNAT for you.

The oVirt Neutron integration in the recent 3.4 release [1] greatly improved
the integration and it now includes IP address management, so you'll be
able to do *nearly* everything from the oVirt GUI.

[1] http://www.ovirt.org/OVirt_3.4.0_release_notes

 I found some workarounds, but no one was easy to implement or the infos was
 not cobering all the aspects.
 
 I have seen the possibility to use a hook, but no user guide and step by
 step instruction found.
 
 Looking forward to get a working solution :-)
 
 Cheers,
 
 -Jarno
 ___
 Users mailing list
 Users@ovirt.org
 http://lists.ovirt.org/mailman/listinfo/users

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] A/B network setup

2014-02-06 Thread Assaf Muller


- Original Message -
 From: Pat Pierson ihasn2...@gmail.com
 To: users@ovirt.org
 Sent: Wednesday, February 5, 2014 10:17:54 PM
 Subject: [Users] A/B network setup
 
 I am having some issues wrapping my head around this but what I am trying to
 setup is a A/B testing environment with a 3node cluster. Each node has 2
 nics, 1 for ovirtmgmt and 1 for vlaned A/B network. I guess what I am trying
 to understand is if ovirt is tagging the vlan's I setup and is properly
 passing that to the switch? 

Yep! You can create two VLAN networks from the GUI (Named A, B for example)
and VLAN tag them with 100 and 200. Then, go to each of your hosts, click
on setup networks and drag those two new networks on the second NIC of the host.
This will configure two VLAN devices on top of the second NIC of the host.
Essentially the NIC will be a trunk port and will allow VLANs 100 and 200.
If the 2nd NIC is called eth1, then after the configuration is done you
will see eth1.100, and eth1.200.

Each network will also have a bridge created for it. So you'll have a Linux
bridge connected to eth1.100, and another one connected to eth1.200. Any
VMs that you connect to the first or second network will have tap devices
connected to the first or second bridge.

 This would allow me to have multiple subnets on
 a host that the switch also can see as multiple vlans. Right? I would still
 have to configure the switch to allow that port access to both vlans.
 

Yeah, you have to make sure that the port connected from the switch to the host
is in trunk mode, and allows VLANs 100 and 200.

 --
 Patrick Pierson
 
 ___
 Users mailing list
 Users@ovirt.org
 http://lists.ovirt.org/mailman/listinfo/users
 
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Networking. Hosted Setup. All in One Host. Hetzner

2014-02-05 Thread Assaf Muller
Could you explain further why does the host need to do any routing?


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Dan Kenigsberg dan...@redhat.com
To: Peter Styk polf...@gmail.com, amul...@redhat.com
Cc: users@ovirt.org
Sent: Wednesday, February 5, 2014 1:23:54 PM
Subject: Re: [Users] Networking. Hosted Setup. All in One Host. Hetzner

On Thu, Jan 16, 2014 at 11:51:25PM +, Peter Styk wrote:
 Greetings,
 
 I'm writing here as to share some of my findings about hosting with
 Hetzner. All in one setups on single remote host can be tricky. Provider
 mounted an extra /29 subnet to the main host but none is routed by default
 and host has to become router itself. At the same time single mistake in
 bridging configuration and lost access results in need for re-bootstrap.
 It's still tempting to try and with many trials I eventually got to see
 guests talking to the net.
 
 Scenario 1: Working. Package bridge-utils, oVirt engine, setup bridge,
 VDSM, add host to engine, add routing to host routing table. Networking by
 trial and error. Still something is not right. Occasionally on ping out I'm
 getting Redirect Host (New nexthop messages.
 http://styk.tv/wp-content/uploads/2014/01/oVirtHosted1_almost_working.png

Unfortunately, I fail to understand what can be hampering your routing
there. Assaf, do you have a guess?

Which version of ovirt have you been using? Now, with source-routing
implemented into ovirt-3.3, there is a danger in setting your own
content into route-netname, as it would be overwritten if netname is
reconfigured via Engine.

 
 Scenario 2: Dreaming. Private network with private router/dhcp/nat. Private
 10.0.0.0/24 network. No problems with routing as gateway 10.0.0.1 would be
 on the same subnet. Thought of using pfSense but can't seem to bring up an
 instance with two network cards on two different networks. I thought this
 would be easy.
 Go to Networks, click create new network, type private, save
 ok. then go to new instance. point at iso, attach two network cards. save
 ok. Launch Host did not satisfy internal filter Network No idea what that
 is. Maybe I don't understand how this works.

I do not understand where having an instance with two nics fail. Is the
Host did not satisfy internal filter Network message coming from
Hetzner management, or oVirt's?

 I even tried removing
 ovirtmgmt network and leaving private network by itself. Tried with all 3
 network card types (rtl8139/e1000/VirtIO)
 http://styk.tv/wp-content/uploads/2014/01/oVirtHosted2_preferred.png

Could you explain how you configured your provate network? In my
experience, your easiest option is to define a dummy interface
ip link add name dummy_private type dummy
and set up a normal oVirt network on top of it, as if it were a true
nic.

 
 Either way if you have a minute or two please take a look at both attached
 diagrams. Deliberately making it difficult by forcing all elements on
 single box in hosted environment.
 
 Maybe there is a way to have this all installed with Neutron or vSwitch on
 the same box or is that pushing it?
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Mixing tagged and untagged VLANs

2014-02-04 Thread Assaf Muller
 Is it not possible to have multiple untagged VLAN networks associated
to one interface in oVirt?

No, not at this time.

You can have one untagged network and N tagged networks on the same device,
but only up to one untagged network.

If you need multiple untagged networks on a single device then you're very
welcome to report an RFE :)


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Trey Dockendorf treyd...@gmail.com
To: users users@ovirt.org
Sent: Monday, February 3, 2014 10:45:44 PM
Subject: [Users] Mixing tagged and untagged VLANs

Using 3.3.2 I seem unable to mix tagged and untagged VLANs on a single
interface.  I'm trying to put the following logical networks on a
host's eth0.

ovirtmgmt:
 - Display Network
 - Migration Network
 - NOT VM Network
 - NO VLAN

private:
 - VM network
 - NO VLAN

ipmi:
 - VM Network
 - VLAN 2

In the host's network setup ovirtmgmt is already linked to eth0.  If I
attach 'ipmi (VLAN 2)' then try and attach 'private' the message is
Cannot have more than one non-VLAN network on one interface.  Same
occurs if I try and attach 'private' when only 'ovirtmgmt' is assigned
to eth0.

Is it not possible to have multiple untagged VLAN networks associated
to one interface in oVirt?

Thanks
- Trey
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Error: The following Network Interfaces were specified more than once:

2014-01-29 Thread Assaf Muller
Please run:
vdsClient -s 0 getVdsCaps

And paste the output.

If this is caused by what I think is causing it, then this bug was fixed
during the 3.4 development cycle and is available in the 3.4 beta release.

(It affects dummy devices and VLANs).

Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Alan Murrell li...@murrell.ca
To: users@ovirt.org
Sent: Wednesday, January 29, 2014 10:27:21 AM
Subject: [Users] Error: The following Network Interfaces were specified
more than   once:

I have noticed what seems to me to be a quirk, but wanted to see if  
this is correct behaviour, or perhaps my configuration.

I am running a test install ov oVirt 3.3.2 in an all-in-one install.   
My test box only has one physical NIC, so I am using that for  
management/external network access, and I have create a dummy  
interface called dummy_0 to be used for the VMs.

I created four networks, each assigned their own VLAN (I am wanted to  
simulate a real-world  .  I added the first network (VLAN001) to  
dummy_0 with no problem (I was able to save the configuration).  The  
system created the dummy_0.1.

When I tried to add another network (VLAN002) to dummy_0 and save  
it, I got the following error:

--- START ---
Error while executing action:

local_host:

 Cannot setup Networks. The following Network Interfaces were  
specified more than once: dummy_0.1.
--- END ---

I found that in the Setup Host Network if I dragged VLAN001 off  
dummy_0 to the right under Required, and left VLAN002 on  
dummy_0, I could save it, and VLAN001 appeared to remain active on  
the host.  Indeed, when I went back into Setup Host Network, both  
VLAN001 and VLAN002 were showing up on dummy_0.

However, if I highlight VLAN002 under Networks and click on the  
Hosts tab, the Network Interface is showing a red down arrow,  
indicating it is not up.  The Network Interface for VLAN001 is a  
green up arrow.  I have refreshed several times and at this point in  
time, have left it for almost thirty minutes, but it remains as a red  
down arrow.  The same is true for the other two networks I tried  
adding: I was able to add them as above, but they are also showing a  
red down arrow.

My questions:

   1.) Is doing what I had to do to add additional networks to an  
interface normal, or is it a quirk of using a dummy interface?
   2.) Any ideas why my additional networks are showing a red down  
arrow?  Again, maybe a quirk of using a dummy interface?

I am going to leave it overnight and see what happens (maybe some  
background process needs to run??), but if that does not work, I am  
sure I can scrounge up another NIC in the next couple of days and  
install it into my test server to see if it does the same thing.

Thanks, in advance.

-Alan
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Error: The following Network Interfaces were specified more than once:

2014-01-29 Thread Assaf Muller
Thanks Alan, that confirms that it was the issue I had in mind.

This bug is fixed in oVirt 3.4, which has a first beta release out.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Alan Murrell li...@murrell.ca
To: users@ovirt.org
Sent: Wednesday, January 29, 2014 4:58:59 PM
Subject: Re: [Users] Error: The following Network Interfaces were  
specified   more than   once:

Quoting Assaf Muller amul...@redhat.com:

 Please run:
 vdsClient -s 0 getVdsCaps

Here you go:

--- START ---
 HBAInventory = {'FC': [], 'iSCSI': [{'InitiatorName':  
'iqn.1994-05.com.redhat:63b5586465eb'}]}
 ISCSIInitiatorName = 'iqn.1994-05.com.redhat:63b5586465eb'
 bondings = {'bond0': {'addr': '',
   'cfg': {},
   'hwaddr': '00:00:00:00:00:00',
   'ipv6addrs': [],
   'mtu': '1500',
   'netmask': '',
   'slaves': []},
 'bond1': {'addr': '',
   'cfg': {},
   'hwaddr': '00:00:00:00:00:00',
   'ipv6addrs': [],
   'mtu': '1500',
   'netmask': '',
   'slaves': []},
 'bond2': {'addr': '',
   'cfg': {},
   'hwaddr': '00:00:00:00:00:00',
   'ipv6addrs': [],
   'mtu': '1500',
   'netmask': '',
   'slaves': []},
 'bond3': {'addr': '',
   'cfg': {},
   'hwaddr': '00:00:00:00:00:00',
   'ipv6addrs': [],
   'mtu': '1500',
   'netmask': '',
   'slaves': []},
 'bond4': {'addr': '',
   'cfg': {},
   'hwaddr': '00:00:00:00:00:00',
   'ipv6addrs': [],
   'mtu': '1500',
   'netmask': '',
   'slaves': []}}
 bridges = {'VLAN001': {'addr': '',
   'cfg': {'DEFROUTE': 'no',
   'DELAY': '0',
   'DEVICE': 'VLAN001',
   'NM_CONTROLLED': 'no',
   'ONBOOT': 'yes',
   'STP': 'no',
   'TYPE': 'Bridge'},
   'gateway': '0.0.0.0',
   'ipv6addrs': ['fe80::2ccd:eff:fee5:6a65/64'],
   'ipv6gateway': '::',
   'mtu': '1500',
   'netmask': '',
   'ports': ['dummy_0.1'],
   'stp': 'off'},
   'VLAN002': {'addr': '',
   'cfg': {'DEFROUTE': 'no',
   'DELAY': '0',
   'DEVICE': 'VLAN002',
   'NM_CONTROLLED': 'no',
   'ONBOOT': 'yes',
   'STP': 'no',
   'TYPE': 'Bridge'},
   'gateway': '0.0.0.0',
   'ipv6addrs': ['fe80::2ccd:eff:fee5:6a65/64'],
   'ipv6gateway': '::',
   'mtu': '1500',
   'netmask': '',
   'ports': ['dummy_0.2'],
   'stp': 'off'},
   'VLAN009': {'addr': '',
   'cfg': {'DEFROUTE': 'no',
   'DELAY': '0',
   'DEVICE': 'VLAN009',
   'NM_CONTROLLED': 'no',
   'ONBOOT': 'yes',
   'STP': 'no',
   'TYPE': 'Bridge'},
   'gateway': '0.0.0.0',
   'ipv6addrs': ['fe80::2ccd:eff:fee5:6a65/64'],
   'ipv6gateway': '::',
   'mtu': '1500',
   'netmask': '',
   'ports': ['dummy_0.9'],
   'stp': 'off'},
   'VLAN010': {'addr': '',
   'cfg': {'DEFROUTE': 'no',
   'DELAY': '0',
   'DEVICE': 'VLAN010',
   'NM_CONTROLLED': 'no',
   'ONBOOT': 'yes',
   'STP': 'no',
   'TYPE': 'Bridge'},
   'gateway': '0.0.0.0',
   'ipv6addrs

Re: [Users] oVirt 3.4 test day summary

2014-01-28 Thread Assaf Muller
The sooner we know the date for the upcoming 2nd test day, the sooner we can
clear that day of meetings.

Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Doron Fediuck dfedi...@redhat.com
To: users users@ovirt.org, bo...@ovirt.org
Sent: Tuesday, January 28, 2014 1:44:40 PM
Subject: [Users] oVirt 3.4 test day summary

Hi all,
thanks for joining us to the first 3.4 test day. We all struggled
with network bandwidth, and hopefully next testday will be better,
so join us again for the next oVirt test day.

Here are some statistics:

Total of bugs opened: 108

Broken into areas:
Unnamed 16
gluster 1
infra 21
integration 8
network 10
node 2
sla 3
storage 24
ux 1
virt 22


Top 3 reporters:
amure...@redhat.com20
emes...@redhat.com7
oourf...@redhat.com7


IRC participants, top 3:
115 sbonazzo
100 didi
81 OaaSvc

Special contributors:
Andrew Lau (and...@andrewklau.com) for his work and reports on hosted engine.

For their special contribution, Andrew Lau and OaaSvc will get special T-shirts
of the oVirt community (please contact me offline).

Thanks again and see you soon in the 2nd test day!

Doron
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] another networking question: ssh to VM with private dhcp addr

2014-01-23 Thread Assaf Muller
I assume when you're creating VMs you connect them to the ovirtmgmt network?

If so, when the VM generates a DHCP request it goes out to the same physical
network that the ovirtmgmt device is on. Meaning, that if the VM and the 
management
device both generate DHCP requests they should get responses in the same subnet.

As for SSH'ing into those VMs, DHCP vs static addressing shouldn't make a 
difference
unless you end up with different subnets.

If you give a VM a static address in the same range as the DHCP responses I'd
imagine you wouldn't be able to SSH into it either.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Robert Story rst...@tislabs.com
To: users users@ovirt.org
Sent: Thursday, January 23, 2014 7:21:01 AM
Subject: [Users] another networking question: ssh to VM with private dhcp   
addr

Hello,

I've got oVirt 3.3.x all-in-one up and running on CentOS 6.3. The IP
address on the ovirtmgmt bridge is static/public. I have not done any
additional network/VLAN configuration.  If I create a VM and give it an
address on the same subnet, I can ssh to it just fine.

If I create a VM and let it use dhcp, it gets a 172.31.x.x address.
Outgoing access works fine, but I can't figure out what to do to allow me
to be able to ssh into those VMs.  With straight KVM the VM private network
was on a virtual interface on the host, making it easy. But I can't figure
out how to do it with oVirt.

Any help greatly appreciated. Thanks in advance.


Robert

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Centos 6.5 and bonding: A slave interface is not properly configured

2014-01-23 Thread Assaf Muller
Hi Karli,

Normally we configure the host's default gateway on the ovirtmgmt network.

If your engine is in the same subnet as all of your hosts, then it's not to 
define
a getway on the ovirtmgmt network.

From 3.3, you can define gateways on the other networks. These will be used via
source routing. For example, if you set a gateway on the display network, but 
leave
the gateway blank for the ovirtmgmt network, then you'll see that 'ip route' 
won't
list a default gateway. However, we configure source routing so that all traffic
coming in to the display network on the host, will be returned via that 
network's
gateway.

You can view source routing rules via:
ip rule

Then ip route show table %s, for each table listed in ip rule.

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: Karli Sjöberg karli.sjob...@slu.se
To: dan...@redhat.com
Cc: amul...@redhat.com, fs...@uncu.edu.ar, users@ovirt.org
Sent: Thursday, January 23, 2014 12:00:52 PM
Subject: Re: [Users] Centos 6.5 and bonding: A slave interface is not properly 
configured

On Thu, 2014-01-23 at 09:34 +, Dan Kenigsberg wrote:
 On Thu, Jan 23, 2014 at 06:50:59AM +, Karli Sjöberg wrote:
  
   
   One comment to your wiki page: in order to better intergrate with our
   recent source-routing support (mostly useful for display network),
   please set GATEWAY only in ifcfg-ovirtmgmt, and remember to set
   DEFROUTE=no in any other ifcfg that mentions IPADDR so as not to
   override the default.
   
   Dan.
  
  Does that imply that default gateway is supposed to be on the same net
  as ovirtmgmt?
 
 Yes, and that's what happening whenever you make changes to
 ovirtmgmt via Vdsm.

To tell you the truth, that makes me rather uncomfortable since we have
set up our ovirtmgmt on a non-routed, black net, only used for _mgmt_
(thought that was the point). The display network is set up on a another
logcal network, a routed net, which is where we also have the default
gateway. Is this now a wrong/bad way to go about it?

/K
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] networking: basic vlan help

2014-01-23 Thread Assaf Muller
If you enable VLAN tagging on the management network, which is configured
on eth0 (Which also provides internet access from my understanding) then
you will connectivity as (I assume) your physical switches aren't configured
for VLANs.

For an all-in-one, what I would suggest is the following procedure:
On your PC, create a dummy NIC via:
sudo ip link add dev dummy_0 type dummy
sudo ip link set dev dummy_0 up

It's important that the name will be in the dummy_* format.

Following that, go back to the GUI, select the host and hit Refresh Host 
Capabilities.

You should see the new dummy_0 device as a host NIC.

Create a VM network, and under the host Network Interfaces tab hit Setup Host 
Networks.

Drag and drop the new VM network on dummy_0 (Don't give dummy_0 a boot protocol 
or an IP address
in the edit network dialog).

At this point you should be able to attach VM vNICs' to the new VM network and 
they won't
be physically connected to any other network, but they'll be able to talk 
amongst themselves.


The private network feature is planned* for oVirt 3.5, so in the future 
you'll be able
to just define a network as a private one and everything will work 
automatically.

* No promises!


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Robert Story rst...@tislabs.com
To: users users@ovirt.org
Sent: Thursday, January 23, 2014 5:44:25 PM
Subject: [Users] networking: basic vlan help

Hello again,

I'm new to VLANs and have a few questions. Right now I just have the mgmt
interface (bridged with eth0) on my all-in-one oVirt test setup. I want to
separate some VMs from the public facing net, which I think means that they
need to be on a different VLAN.  I created two new networks, pubX and
privY, with vlan ids X and Y, but couldn't assign them to eth0 because the
current mgmt network is non-VLAN. I was about to enable VLAN tagging on the
mgmt network, but I wanted to make sure that doing so wouldn't do anything
to eth0 that would disrupt access to it (I only have remote access and don't
want to lock myself out).  Also, if it is safe, does the mgmt vlan tag id
matter? is 0 the right value?

Any/all help, hints, tips or references to examples/links greatly
appreciated.


Robert

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] networking: basic vlan help

2014-01-23 Thread Assaf Muller
Sorry, privY on eth1.

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: Robert Story rst...@tislabs.com
To: Assaf Muller amul...@redhat.com
Cc: users users@ovirt.org
Sent: Thursday, January 23, 2014 7:11:50 PM
Subject: Re: [Users] networking: basic vlan help

On Thu, 23 Jan 2014 10:59:57 -0500 (EST) Assaf wrote:
AM If you enable VLAN tagging on the management network, which is
AM configured on eth0 (Which also provides internet access from my
AM understanding) then you will connectivity as (I assume) your physical
AM switches aren't configured for VLANs.

I'm assuming will connectivity should have been will lose connectivity,
which is what I feared. I'm glad I asked!

AM For an all-in-one, what I would suggest is the following procedure:

Excellent, I'll try that. Thanks!

My next question is for future planning. There is a second interface
(eth1) with a separate physical network which only contains the engine,
nodes and the nfs server. 

 +--+
 | internet |-|---|--|
 +--+ ++  +---+  +---+   eth0
  | engine |  | node1 |  | node2 |
+-+   ++  +---+  +---+   eth1
| nfs |---|---|--|
+-+

Can the mgmt network be easily moved to eth1? Then the pubX would be
non-vlan on eth0, and mgmt + privY would be on eth1. If all the eth1
interfaces are connected to a dedicated/isolated switch, does that switch
need to explicitly support vlans, or does it matter?



Robert

--
Senior Software Engineer @ Parsons
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] networking: basic vlan help

2014-01-23 Thread Assaf Muller
What is the purpose of PubY on eth1?

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: Robert Story rst...@tislabs.com
To: Assaf Muller amul...@redhat.com
Cc: users users@ovirt.org
Sent: Thursday, January 23, 2014 7:11:50 PM
Subject: Re: [Users] networking: basic vlan help

On Thu, 23 Jan 2014 10:59:57 -0500 (EST) Assaf wrote:
AM If you enable VLAN tagging on the management network, which is
AM configured on eth0 (Which also provides internet access from my
AM understanding) then you will connectivity as (I assume) your physical
AM switches aren't configured for VLANs.

I'm assuming will connectivity should have been will lose connectivity,
which is what I feared. I'm glad I asked!

AM For an all-in-one, what I would suggest is the following procedure:

Excellent, I'll try that. Thanks!

My next question is for future planning. There is a second interface
(eth1) with a separate physical network which only contains the engine,
nodes and the nfs server. 

 +--+
 | internet |-|---|--|
 +--+ ++  +---+  +---+   eth0
  | engine |  | node1 |  | node2 |
+-+   ++  +---+  +---+   eth1
| nfs |---|---|--|
+-+

Can the mgmt network be easily moved to eth1? Then the pubX would be
non-vlan on eth0, and mgmt + privY would be on eth1. If all the eth1
interfaces are connected to a dedicated/isolated switch, does that switch
need to explicitly support vlans, or does it matter?



Robert

--
Senior Software Engineer @ Parsons
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] networking: basic vlan help

2014-01-23 Thread Assaf Muller
Then the currently offered topology is to have ovirtmgmt on eth1 untagged, and 
non-VM, and have
privY as a VM, tagged network on eth1. That would require the physical switch 
to be VLAN-aware
and configured properly. oVirt wise you should have no problems using the GUI 
to move to that
topology once you've decided to do so.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Robert Story rst...@tislabs.com
To: Assaf Muller amul...@redhat.com
Cc: users users@ovirt.org
Sent: Thursday, January 23, 2014 8:41:43 PM
Subject: Re: [Users] networking: basic vlan help

On Thu, 23 Jan 2014 13:33:07 -0500 (EST) Assaf wrote:
AM Sorry, privY on eth1.

For VM to VM communication that doesn't need to go over the public net..


Robert

--
Senior Software Engineer @ Parsons
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Unbundle bond0 which virtmgmt is attached

2014-01-19 Thread Assaf Muller
There's two checkboxes in the setup networks dialog:

1) save network configuration - This only means that if the change was 
successful, persist the final networking state
   so that when the host is rebooted, it is booted into the network state 
*after* the (successful) networking change
2) check connctivity - After the network operation, verify that the engine 
still has connectivity to the host. If so,
   commit the change. Otherwise roll it back.

The save network configuration check box shouldn't change the outcome of the 
result.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: William Kwan pota...@yahoo.com
To: Dan Kenigsberg dan...@redhat.com
Cc: users@ovirt.org
Sent: Sunday, January 19, 2014 4:49:15 PM
Subject: Re: [Users] Unbundle bond0 which virtmgmt is attached

Thanks a lot! It is a good pointer, it is working. 

I'm not sure this is a dumb error from a new user. When I change the network 
configuration on the first system, I checked the save network configuration. 
It didnt' work and dropped the connection to the system. 

1. I had both systems in maintenance mode 
2. from hosts, picked one of the system, setup host networks 
3. Right clicked on eth1, Remove eth0 from Bond 
4. at this point, ovirtmgmt was on the right hand side listed as Unassigned 
Logical networks 
5. I dragged it back to assign it to eth0 
6. checked save network configuration 
6. clicked OK and connection is dropped 

Odd... ON the console, I saw the followings 
bonding: unable to remove non-existent slave eth0 for bond bond0. 
bonding: unable to remove non-existent slave eth1 for bond bond0. 

`/etc/init.d/network restart` and I got bond0 back. I redo the whole process 
without save network configuration and it worked fine. I did this on the 2nd 
system and it worked fine also. 


Will 





On Saturday, January 18, 2014 7:56 PM, Dan Kenigsberg dan...@redhat.com 
wrote: 
On Sat, Jan 18, 2014 at 03:17:22PM -0800, William Kwan wrote: 

 Hi, 
 
 FAQ type of question. 
 
 If virtmgmt is assigned to bond0, which is composed of eth0 and eth1, what is 
 the best way to unbundle eth1 from bond0? 


I'd say: 
- Move the host to maintanence, so it runs no VMs and does not play SPM 
role. 
- Use the Setup Network dialog to brake the bond and define ovirtmgmt 
directly on top of eth0. The Engine/Vdsm connection should survive 
this kind of change (unlike a change of management address or vlan). 

If the host uses a special storage domain, or is not SPM, you could even 
do that while VMs are running. 





___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] vlan on mgmt network

2014-01-16 Thread Assaf Muller
You can make ovirtmgmt a non-VM, untagged (non-VLAN) network, then place
as many VM, tagged networks as you'd like on the same NIC or bond.

Here's a working upgrade procedure:
1) Edit ovirtmgmt, mark it as non-VM.
2) Go to all hosts using this network, select Setup Host Networks. You'll see 
that
   ovirtmgmt is out of sync. Edit the network and mark sync. Hit ok on the 
dialog of course :)
   Step 2 will be done for you automatically when you edit logical networks in 
oVirt 3.4
3) Create VM, VLAN networks
4) Go to all hosts you wish to place those networks, and drag the new VLAN 
networks on
   bond0.

You don't need to do any prep work on your hosts before this procedure. oVirt 
will take care
of this stuff for you.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Christian Hernandez christi...@4over.com
To: William Kwan pota...@yahoo.com
Cc: users@ovirt.org
Sent: Thursday, January 16, 2014 6:36:49 PM
Subject: Re: [Users] vlan on mgmt network

Assuming it's on the same VLAN... 

I would make the subinterface (i.e. bond0.1) 


Thank you, 

Christian Hernandez 
1225 Los Angeles Street 
Glendale, CA 91204 
Phone: 877-782-2737 ext. 4566 
Fax: 818-265-3152 
christi...@4over.com mailto: christi...@4over.com  
www.4over.com  http://www.4over.com  


On Thu, Jan 16, 2014 at 8:31 AM, William Kwan  pota...@yahoo.com  wrote: 



Hi all, 

CentOS 6.5 and oVirt 3.3 

Let's say I have bond0 (eth0 and eth1) on a management network. 
Logical network ovirtmgmt is set and working on top of Interfaces bond0. 

Is it possible to have another Logical network, e.g. ovirt777, on the same 
Interface bond0? 

Or I should add a virtual interface, e.g. bond0.1, tie it to VLAN 777 and 
create logical network with this virtual interface? 

Wil 


___ 
Users mailing list 
Users@ovirt.org 
http://lists.ovirt.org/mailman/listinfo/users 



___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] vlan on mgmt network

2014-01-16 Thread Assaf Muller
Ok so if you want to set the management network (untagged) and VM networks 
(tagged) on the same bond,
then the management network has to be non-VM. IE: You won't be able to connect 
instances to the management
network.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: William Kwan pota...@yahoo.com
To: Assaf Muller amul...@redhat.com
Cc: users@ovirt.org, Christian Hernandez christi...@4over.com
Sent: Thursday, January 16, 2014 7:59:46 PM
Subject: Re: [Users] vlan on mgmt network

Assaf,

Thanks a lot.   Will I be able to mix tagged and untagged traffic?

Example:
- bond0 on my two systems are on the management network, untagged
- my VMs all have one connection to this management network
- I want to setup a new VM with two interfaces, one on the management network, 
one on the tagged VLAN


This will work it out?
Thanks
Will




On Thursday, January 16, 2014 12:26 PM, Assaf Muller amul...@redhat.com wrote:
 
You can make ovirtmgmt a non-VM, untagged (non-VLAN) network, then place
as many VM, tagged networks as you'd like on the same NIC or bond.

Here's a working upgrade procedure:
1) Edit ovirtmgmt, mark it as non-VM.
2) Go to all hosts using this network, select Setup Host Networks. You'll see 
that
   ovirtmgmt is out of sync. Edit the network and mark sync. Hit ok on the 
dialog of course :)
   Step 2 will be done for you automatically when you edit logical networks in 
oVirt 3.4
3) Create VM, VLAN networks
4) Go to all hosts you wish to place those networks, and drag the new VLAN 
networks on
   bond0.

You don't need to do any prep work on your hosts before this procedure. oVirt 
will take care
of this stuff for you.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Christian Hernandez christi...@4over.com
To: William Kwan pota...@yahoo.com
Cc: users@ovirt.org
Sent: Thursday, January 16, 2014 6:36:49 PM
Subject: Re: [Users] vlan on mgmt network

Assuming it's on the same VLAN... 

I would make the subinterface (i.e. bond0.1) 


Thank you, 

Christian Hernandez 
1225 Los Angeles Street 
Glendale, CA 91204 
Phone: 877-782-2737 ext. 4566 
Fax: 818-265-3152 
christi...@4over.com mailto: christi...@4over.com  
www.4over.com  http://www.4over.com  


On Thu, Jan 16, 2014 at 8:31 AM, William Kwan  pota...@yahoo.com  wrote: 



Hi all, 

CentOS 6.5 and oVirt 3.3 

Let's say I have bond0 (eth0 and eth1) on a management network. 
Logical network ovirtmgmt is set and working on top of Interfaces bond0. 

Is it possible to have another Logical network, e.g. ovirt777, on the same 
Interface bond0? 

Or I should add a virtual interface, e.g. bond0.1, tie it to VLAN 777 and 
create logical network with this virtual interface? 

Wil 


___ 
Users mailing list 
Users@ovirt.org 
http://lists.ovirt.org/mailman/listinfo/users 




___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] no VM network connection

2014-01-10 Thread Assaf Muller
If you intend to produce data following your stress tests, please share.

Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: William Kwan pota...@yahoo.com
To: Dan Kenigsberg dan...@redhat.com, Assaf Muller amul...@redhat.com
Cc: users@ovirt.org
Sent: Thursday, January 9, 2014 11:26:50 PM
Subject: Re: [Users] no VM network connection

Thanks for the tips. Yes, after tcpdump left and right, we found an issue.  It 
is definitely an user error.

BTW, thanks for all the assistance.  We have a promising setup working so far.  
We are running tests and putting more stress to test it.

Thanks
Will




On Thursday, January 9, 2014 11:53 AM, Dan Kenigsberg dan...@redhat.com wrote:
 
On Thu, Jan 09, 2014 at 09:18:15AM -0500, Assaf Muller wrote:
 Hi Will,
 
 Do these new faulty VMs get IP from the DHCP server on subnet1? If so, can 
 they ping their default gateway?
 
 If you SSH into the hypervisor that hosts the new VM, do you see a new 
 'vnetX' device for it? If you type brctl showmacs name of VM bridge, do you 
 see the MAC of the NIC of the VM that is connected to the VM network?
 
 If you tcpdump bond1, do you see any outgoing VM traffic?

Also, if you set an explicit IP address within the guest and try to ping
outside of it - cun you see packets travelling out of the vnetX and into
the connected bridge?


 
 Assaf Muller, Cloud Networking Engineer 
 Red Hat 
 
 
 - Original Message -
 From: William Kwan pota...@yahoo.com
 To: users@ovirt.org
 Sent: Wednesday, January 8, 2014 8:47:51 PM
 Subject: [Users] no VM network connection
 
 Hi all, 
 
 Need some pointers here. I created one VM successfully with network 
 connection. New VMs and clone can't seem to get network connections. 
 
 I have ovirt 3.3.2-1.el6 on CentOS6.5. Two virtualization nodes are on 6.5 
 with GlusterFS. 
 Both nodes have bond0 on management subnet and bond1 on subnet1. 
 bond0 has an IP assigned and bond1 has no IP assigned (just link up) 
 
 I created a VM with two NICs attached to two the logical networks. This VM1 
 works fine with both network connections. This VM can be started on either 
 node fine with network connections. 
 
 When I tried to build a new VM or build one from the template of VM1, I can't 
 seem to get another network connection working on these new VMs. Yes, they 
 have logical network attached to them.
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Network unsynchronised

2014-01-09 Thread Assaf Muller
If you check out /var/log/vdsm/supervdsm.log, do you see any setupNetworks 
being called as a result of you hitting sync network?

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: Alan Murrell li...@murrell.ca
To: users@ovirt.org
Sent: Thursday, January 9, 2014 10:13:27 AM
Subject: [Users] Network unsynchronised

Network unsynchronised 

Hi All. 

I have a brand new, fresh all-in-one installation that I am using to evaluate 
oVirt. 

I have removed the VM network role from the ovirtmgmt bridge, as I want to 
use it for management only, and no VMs (I plan on using other networks I am 
creating with VLAN IDs for handling the VM network). I can remove the role 
alright, but when I go in to Setup Host Network, the ovirtmgmt network is 
showing as being unsynchronised. When I click on the pencil icon on the network 
to edit it, I put a check in Sync Network and clik OK but it never seems to 
sync. 

Any idea why it will not? 

Thanks! :-) 

-Alan 


___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] no VM network connection

2014-01-09 Thread Assaf Muller
Hi Will,

Do these new faulty VMs get IP from the DHCP server on subnet1? If so, can they 
ping their default gateway?

If you SSH into the hypervisor that hosts the new VM, do you see a new 'vnetX' 
device for it? If you type brctl showmacs name of VM bridge, do you see the 
MAC of the NIC of the VM that is connected to the VM network?

If you tcpdump bond1, do you see any outgoing VM traffic?

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: William Kwan pota...@yahoo.com
To: users@ovirt.org
Sent: Wednesday, January 8, 2014 8:47:51 PM
Subject: [Users] no VM network connection

Hi all, 

Need some pointers here. I created one VM successfully with network connection. 
New VMs and clone can't seem to get network connections. 

I have ovirt 3.3.2-1.el6 on CentOS6.5. Two virtualization nodes are on 6.5 with 
GlusterFS. 
Both nodes have bond0 on management subnet and bond1 on subnet1. 
bond0 has an IP assigned and bond1 has no IP assigned (just link up) 

I created a VM with two NICs attached to two the logical networks. This VM1 
works fine with both network connections. This VM can be started on either node 
fine with network connections. 

When I tried to build a new VM or build one from the template of VM1, I can't 
seem to get another network connection working on these new VMs. Yes, they have 
logical network attached to them. 

Thanks in advance, 

Will 



___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Another issue - VM Networking

2014-01-05 Thread Assaf Muller
I haven't run into this particular problem, but we need more information.

What is the networking like on each node, the same I imagine? I understand
it's VLAN - bridge - nic?

Can we grab /var/log/vdsm/vdsm.log, and supervdsm.log during the start of a VM,
and then during the migration to the faulty node?
(You can paste the full contents of those logs to pastebin).

Are you guys using any hooks?


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Dan Ferris dfer...@prometheusresearch.com
To: users@ovirt.org
Sent: Saturday, January 4, 2014 1:13:36 AM
Subject: [Users] Another issue - VM Networking

I've run into another issue with Ovirt 3.3.2.

We have a 2 node VM cluster with Fibre Channel storage.  After upgrade
everything to FC 19 and installing Ovirt 3.3 from scratch one of the
nodes has network issues.

The issue is that when we start a VM on the node, the VM cannot access
the network.  No ping, no ssh, nothing.  However, if I try to ping the
VM, I can see the ICMP traffic on the ethernet VLAN interface and the
bridge interface.  I cannot see any traffic on the VM tap interface.

If we migrate the VM from the faulty node to the other node that works,
and then back, the network will start to work.  However, if the VM is
shut down the network will fail again when it's booted on the faulty node.

Does anyone have any ideas about this particular problem?

Thanks,

Dan
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Network issues - Bonding

2014-01-05 Thread Assaf Muller
If you ifdown bond4 then ifup it, does the bond mode properly update
to bond mode 1? If not, it sounds like an initscripts or bonding module bug.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Dan Ferris dfer...@prometheusresearch.com
To: users users@ovirt.org
Sent: Saturday, January 4, 2014 5:31:21 AM
Subject: [Users] Network issues - Bonding

Hello All,

A little bit ago I wrote an email about network issues I was having.

I found the problem...

On the VM host, I had a bond set up between two network interfaces.  The
bond mode was set to mode 1 (active/passive).

However when I look at the bond on the box, I get this:

[root@node02 bonding]# cat bond4
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: load balancing (round-robin)
MII Status: up
MII Polling Interval (ms): 0
Up Delay (ms): 0
Down Delay (ms): 0

Slave Interface: em2
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: d4:ae:52:6d:c8:cc
Slave queue ID: 0

Slave Interface: em3
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: d4:ae:52:6d:c8:ce
Slave queue ID: 0

Somehow, the OS is not setting the bonding mode right.  I verified that
it was set to mode 1 in /etc/sysconfig/network-scripts/ifcfe-bond4

When I take the bond away, the host network works perfectly on both of
the formerly bonded interfaces.

So again, if anyone has any ideas, I'm open to suggestions.

Thanks,

Dan
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Virtual networks

2014-01-05 Thread Assaf Muller
oVirt currently cannot, however with the oVirt / Neutron integration* you'll be 
able to do that in the upcoming 3.4 release.

You can read more at:
http://www.ovirt.org/Features/OSN_Integration
http://www.ovirt.org/Features/Detailed_OSN_Integration
http://www.ovirt.org/images/e/e7/Ovirt-neutron-integration-deep-dive-2013.pdf

* The oVirt / Neutron integration is in tech preview mode and is already 
available for 3.3, however VM IP address
  management will be added in 3.4.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Maurice James midnightst...@msn.com
To: users@ovirt.org
Sent: Saturday, January 4, 2014 5:24:52 PM
Subject: [Users] Virtual networks



Can ovirt provide DHCP addressing to VMs the way libvirt does on its own? 

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Error on Setting up VLAN

2014-01-01 Thread Assaf Muller
Looks like bridge public_10 still has vNICs connected to it.

What is the output for:
brctl show public_10

If you select the host and enter the Setup Networks dialog, does the physical 
NIC
have a network on it? If so, remove it (I know you said you did this but it 
looks like
that operation somehow silently failed). After removing all networks on the 
host,
the public_10 bridge shouldn't exist. Also, virsh net-list shouldn't list the 
public_10 network.

If setupNetworks shows a blank NIC, but the public_10 bridge or libvirt 
networks exist, try:
Removing /etc/sysconfig/network-scripts/ifcfg-public_10
Type virsh enter, then:
virsh net-undefine public_10
virsh net-destroy public_10
Login details:
vdsm@ovirt
shibboleth

After the ifcfg file and libvirt networks are removed, go back to the GUI and 
select 'Save network config',
then restart the host.

At this point the host should come up - No ifcfg file for the network, no 
libvirt network, no bridge.
The setup networks dialog in the GUI should show no network on the physical NIC.
Now make sure (again) that the public_10 network is a VLAN network, and attach 
public_10 to the NIC.
That operation should succeed.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Alessandro Bianchi a.bian...@skynet.it
To: users@ovirt.org
Sent: Wednesday, January 1, 2014 7:01:17 PM
Subject: [Users] Error on Setting up VLAN

Hi everyone and happy new year first !

I'm setting up a test host using VLAN

It holds 2 VM's so I

- Stopped the two VMs and removed the public_10 network from the NIC
- edited the public_10 network, checked enable VLAN tagging and set 
10 as VLAN identifier
- assigned the new edited network to the VMs NIC
- edited the HOST network, clicked the pencil on the relevant NIC and 
clicked sync network since it showed the network as not synced

Ovirt tried to setup the network but ended with cannot up interface 
error and I lost contact with the node

Ovirt is 3.3.2 on Fedora 19 (it is AIO)

Here is the log:

53244 Thread-52989::DEBUG::2014-01-01 
16:58:16,938::BindingXMLRPC::974::vds::(wrapper) client 
[172.16.0.5]::call setupNetworks with ({'public_10': {'nic': 'p5p1', 
'vlan': '10', 'ipaddr': '10.0.0.5', 'netmask': '255.255.255.0', 'STP': 
'no', 'bridged': 'true', 'gateway': '10.0.0.254'}}, {}, 
{'connectivityCheck': 'true', 'connectivityTimeout': 120}) {} flowID 
[49756c97]
53245 Thread-52925::DEBUG::2014-01-01 
16:56:39,737::BindingXMLRPC::974::vds::(wrapper) client 
[172.16.0.5]::call ping with () {} flowID [64a8776]
   53246 Thread-52925::DEBUG::2014-01-01 
16:56:39,738::BindingXMLRPC::981::vds::(wrapper) return ping with 
{'status': {'message': 'Done', 'code': 0}}
   53247 Thread-52924::ERROR::2014-01-01 
16:56:39,802::API::1262::vds::(setupNetworks) bridge public_10 has 
interfaces set(['vnet0', 'vnet2']) connected
   53248 Traceback (most recent call last):
   53249   File /usr/share/vdsm/API.py, line 1260, in setupNetworks
   53250 supervdsm.getProxy().setupNetworks(networks, bondings, options)
   53251   File /usr/share/vdsm/supervdsm.py, line 50, in __call__
   53252 return callMethod()
   53253   File /usr/share/vdsm/supervdsm.py, line 48, in lambda
   53254 **kwargs)
   53255   File string, line 2, in setupNetworks
   53256   File /usr/lib64/python2.7/multiprocessing/managers.py, line 
773, in _callmethod
   53257 raise convert_to_error(kind, result)
   53258 ConfigNetworkError: (28, bridge public_10 has interfaces 
set(['vnet0', 'vnet2']) connected)
   53259 Thread-52924::DEBUG::2014-01-01 
16:56:39,808::BindingXMLRPC::981::vds::(wrapper) return setupNetworks 
with {'status': {'message': bridge public_10 has interfaces 
set(['vnet0', 'vnet2']) connected, 'code': 28}}

ifcfg-p5p1 (before loosing connection)
[root@hypervisor ~]# cat /etc/sysconfig/network-scripts/ifcfg-p5p1
DEVICE=p5p1
ONBOOT=yes
HWADDR=00:50:fc:85:9b:76
BRIDGE=public_10
NM_CONTROLLED=no
STP=no


ifcfg-public_10
DEVICE=public_10
ONBOOT=yes
TYPE=Bridge
DELAY=0
NM_CONTROLLED=no
STP=no
IPADDR=10.0.0.5
NETMASK=255.255.255.0
DEFROUTE=yes
GATEWAY=10.0.0.254
PEERDNS=yes
DNS1=8.8.8.8
DNS2=8.8.4.4

How may I amend the configuration files?

Any hint?

Thank you

Best regards


___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] disable EnableMACAntiSpoofingFilterRules per VM

2013-12-26 Thread Assaf Muller
Yep!

Here you go:
'yum install vdsm-hook-macspoof' on all hosts, then following the instructions 
here:
https://github.com/oVirt/vdsm/blob/master/vdsm_hooks/macspoof/README

You can disable the filter on a VM or VNIC level.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Kapetanakis Giannis bil...@edu.physics.uoc.gr
To: users@oVirt.org users@ovirt.org
Sent: Wednesday, December 25, 2013 1:19:23 PM
Subject: [Users] disable EnableMACAntiSpoofingFilterRules per VM

Hi,

Trying to move a VM with CARP/VRRP on oVirt 3.3.2 I got bitten by this.
http://www.ovirt.org/Features/Design/Network/NetworkFiltering
http://lists.ovirt.org/pipermail/users/2013-October/017217.html

Is there a way to disable mac filtering only for a specific VM
and not for the entire cluster?

I've tried giving MAC addresses in the form of
00-00-5E-00-01-XX
but it didn't work.

Best regards,

G
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] disable EnableMACAntiSpoofingFilterRules per VM

2013-12-26 Thread Assaf Muller
 what will be the default action for my VMs? Filter or not filter?
 So I gave to alter EnableMACAntiSpoofingFilterRules as well?

The default is to filter. You'll have to add a VM or VNIC custom property
and set the value to False whenever you want to disable the filtering.

 What about conflicting values between VM and VNIC. Which has precedence 
 over the other?

You would think that anything VNIC specific would take precedence, but with
how the code is implemented at this time, before_device_create is called
for all devices, and before_vm_create is called after that. That means
that whatever is defined at the VM level will take precedence.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Kapetanakis Giannis bil...@edu.physics.uoc.gr
To: users@oVirt.org users@ovirt.org
Sent: Wednesday, December 25, 2013 5:44:47 PM
Subject: Re: [Users] disable EnableMACAntiSpoofingFilterRules per VM

On 25/12/13 17:26, Assaf Muller wrote:
 Yep!

 Here you go:
 'yum install vdsm-hook-macspoof' on all hosts, then following the 
 instructions here:
 https://github.com/oVirt/vdsm/blob/master/vdsm_hooks/macspoof/README

 You can disable the filter on a VM or VNIC level.


 Assaf Muller, Cloud Networking Engineer
 Red Hat

Thanks for the quick reply.

That looks very cool :)

Just to understand better because the instructions are not clear on that.

Suppose I install the package and add
engine-config -s 
UserDefinedVMProperties='previousProperties;macspoof=^(true|false)$' 
--cver=3.3

what will be the default action for my VMs? Filter or not filter?
So I gave to alter EnableMACAntiSpoofingFilterRules as well?

I mean do I have to explicitly define macspoof=true on all the VMs 
except the VMs I don't need filtering
or the opposite - filter by default and set macspoof=false in the VM I 
don't need filtering?

I guess the same applies for VNIC.

What about conflicting values between VM and VNIC. Which has precedence 
over the other?

Thanks

G

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] disable EnableMACAntiSpoofingFilterRules per VM

2013-12-26 Thread Assaf Muller
Thank you for your feedback. I sent a patch to fix all mentioned issues:
http://gerrit.ovirt.org/#/c/22760/

It will be available for oVirt 3.4.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Kapetanakis Giannis bil...@edu.physics.uoc.gr
To: users@oVirt.org users@ovirt.org
Cc: Assaf Muller amul...@redhat.com, dan...@redhat.com
Sent: Thursday, December 26, 2013 3:23:07 PM
Subject: Re: [Users] disable EnableMACAntiSpoofingFilterRules per VM

Resending cause somehow the format got screwed up.

On 26/12/13 11:06, Assaf Muller wrote:
 what will be the default action for my VMs? Filter or not filter?
 So I gave to alter EnableMACAntiSpoofingFilterRules as well?
 The default is to filter. You'll have to add a VM or VNIC custom property
 and set the value to False whenever you want to disable the filtering.

 What about conflicting values between VM and VNIC. Which has precedence
 over the other?
 You would think that anything VNIC specific would take precedence, but with
 how the code is implemented at this time, before_device_create is called
 for all devices, and before_vm_create is called after that. That means
 that whatever is defined at the VM level will take precedence.

ok, this worked like a charm but I did some changes.

Maybe you want to update the README on
https://github.com/oVirt/vdsm/blob/master/vdsm_hooks/macspoof/README

engine-config -s 
UserDefinedVMProperties='previousPropertiesmacspoof=^(true|false)$' 
--cver=3.3

failed with an error:

Cannot set value previousProperties;macspoof=^(true|false)$ to key 
UserDefinedVMProperties. Invalid syntax, user defined VM properties 
specification should conform to 
(([a-z_A-Z0-9])+)=(([^;])+)(;(([a-z_A-Z0-9])+)=(([^;])+))*;?

I changed that to
engine-config -s UserDefinedVMProperties=macspoof=(true|false)
like it is defined in http://www.ovirt.org/Engine_config_examples
which was inserted with no error.

CustomDeviceProperties also inserted like
engine-config -s CustomDeviceProperties=
'{type=interface;prop={ifacemacspoof=(true|false)}}'

so both true/false options could be available.

VM option indeed takes precedence over VNIC option.

You should also take notice that setting macspoof=false or 
ifacemacspoof=false does the opposite of at least what I expected.
It filters.

I had to apply with either macspoof=true or ifacemacspoof=true to 
disable filtering.

The README also has that the other way.
Either the README has to be changed or the options could be renamed to 
something like disable_macspoof_filter and disable_ifacemacspoof_filter

best regards and thanks for all the replies,

G

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] simple networking? [SOLVED] mostly

2013-12-19 Thread Assaf Muller
Juan - Are you using an all-in-one?

Assaf Muller, Cloud Networking Engineer 
Red Hat 


- Original Message -
From: Juan Pablo Lorier jplor...@gmail.com
To: Antoni Segura Puimedon asegu...@redhat.com, Itamar Heim 
ih...@redhat.com
Cc: users@ovirt.org
Sent: Monday, December 16, 2013 9:16:13 PM
Subject: Re: [Users] simple networking? [SOLVED] mostly

Hi Antoni,

I've tried two secentarios: to create a ifcfg-eth0.128 with the
parameters so it survives restarts, and to create dinamically with
vconfig and ifconfig so it does not survive the restart to see if ovirt
get the mgmt working.
The only thing that worked to me was creating the hole ovirtmgmt bridge
on top or the bondX.128 and of course, it does not fails.

This log is from one of the last two hosts I set up (3.3.1) where I
didn't create the bond, just created a vlan on top of eth0.

https://drive.google.com/file/d/0B9OrU8RK9m26TjdGcTBLbkJrYVU/edit?usp=sharing

Regards,


On 16/12/13 16:52, Antoni Segura Puimedon wrote:
 Hi Juan Pablo,

 Could you please share:

 /var/log/vdsm/supervdsm.log

 So that I can see why the installation fails? I'd also like to know
 how you set up the pre-existent vlan on the host.

 - Original Message -
 From: Itamar Heim ih...@redhat.com
 To: Juan Pablo Lorier jplor...@gmail.com, users@ovirt.org, Dan 
 Kenigsberg dan...@redhat.com, Antoni Segura
 Puimedon asegu...@redhat.com
 Sent: Monday, December 16, 2013 7:13:04 PM
 Subject: Re: [Users] simple networking? [SOLVED] mostly

 On 12/16/2013 12:54 PM, Juan Pablo Lorier wrote:
 Itamar,

 I have the same problem and already have mentioned that in previous
 posts. In my case, I have vlan 128 tagged assigned for ovirtmgmt so I
 set the host ip manually on top of a vlan interface and the, after ovirt
 does install the node, it fails to finish as it can't manage properly to
 create the bridge and pass the ip to the brand new bridge. The ovirtmgmt
 interface I create in engine is a bond (and originally I configure just
 a single interface with the ip to grant engine access), just in case
 this changes the scenario.
 Regards,

 ___
 Users mailing list
 Users@ovirt.org
 http://lists.ovirt.org/mailman/listinfo/users

 dan/toni - thoughts?


___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Bridge w/o IP address?

2013-12-19 Thread Assaf Muller
Host A has two NICs.
Setup ovirtmgmt on one NIC, and the internal network on the second NIC.
The internal network should be a VM network with no IP configuration.
If the VMs connected to the internal network will be allocated IPs
from a different subnet than the VMs on the ovirtmgmt network, then
the internal VMs won't be able to ping the host or the ovirtmgmt VMs.

You can also use VLANs to segregate the two networks, which will also
allow you to place both bridges on the same physical network device.
You could bond two NICs, then place two VLANned, VM networks on the same
bond.


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Blaster blas...@556nato.com
To: Users@ovirt.org List users@ovirt.org
Sent: Thursday, December 19, 2013 7:26:17 PM
Subject: [Users] Bridge w/o IP address?

I want to put multiple VMs on the same hypervisor, some will be Internet 
facing, some will be internal facing.

I obviously don't want the hypervisor to be an attack vector, so I don't 
want the HV to have an IP address on the DMZ network.

Is this configuration as simple as creating another another bridge like 
the ovirtmgmt bridge, not giving it an IP address, and adding it to the 
DMZ facing hosts?



___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] ovirtmgmt not installed

2013-12-04 Thread Assaf Muller
I personally have a F19 host working against 3.3 engine. Could you detail what 
doesn't work, and possibly attach engine and VDSM logs when an operation fails?

Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Pascal Jakobi pjak...@yahoo.fr
To: users@ovirt.org
Sent: Wednesday, December 4, 2013 8:22:05 AM
Subject: Re: [Users] ovirtmgmt not installed

At the end, I could not make an F19 Fedora host work against a 3.3 engine. I 
suggest someone from the eng team has a look at it. some stuff just does not 
work there. 

Therefore, I downgraded the fedora host to F18. On the engine side, I got the 
following error message : 
Host lab2 is compatible with versions (3.0,3.1,3.2) and cannot join Cluster 
Default which is set to version 3.3. 
Don't know what is to be set 

Anyway, I created a 3.2 cluster and installed the host into it. Seems to work 
(the host is up). 

Thanks again for support. 


Pascal Jakobi 
116 rue de Stalingrad 
93100 Montreuil, France 
Tel : + 33 6 87 47 58 19 
@ : pjak...@yahoo.fr 


Le Mardi 3 décembre 2013 14h04, Pascal Jakobi pjak...@yahoo.fr a écrit : 
Will do within a couple of hours. 
Many, many thanks 
Pascal Jakobi 
116 rue de Stalingrad 
93100 Montreuil, France 
Tel : + 33 6 87 47 58 19 
@ : pjak...@yahoo.fr 


Le Mardi 3 décembre 2013 13h43, Antoni Segura Puimedon asegu...@redhat.com a 
écrit : 
Hi Pacal, 

Would you mind joining #ovirt at irc.oftc.net and pinging apuimedo? 
I have some questions that could be faster solved online. 

Best, 

Antoni 

- Original Message - 
 From: Pascal Jakobi  pjak...@yahoo.fr  
 To: Pascal Jakobi  pascal.jak...@gmail.com , Moti Asayag  
 masa...@redhat.com  
 Cc: users@ovirt.org 
 Sent: Tuesday, December 3, 2013 1:11:36 PM 
 Subject: Re: [Users] ovirtmgmt not installed 
 
 Still the same, even with networking working like a charm... 
 Any idea ? 
 
 [ root@lab2 vdsm]# vdsClient -s 0 getVdsCaps 
 HBAInventory = {'FC': [], 'iSCSI': [{'InitiatorName': 
 'iqn.1994-05.com.redhat:a0bc8a5f9a5d'}]} 
 ISCSIInitiatorName = 'iqn.1994-05.com.redhat:a0bc8a5f9a5d' 
 bondings = {'bond0': {'addr': '', 
 'cfg': {}, 
 'hwaddr': '96:9d:8f:27:e1:8f', 
 'ipv6addrs': [], 
 'mtu': '1500', 
 'netmask': '', 
 'slaves': []}} 
 bridges = {} 
 clusterLevels = ['3.0', '3.1', '3.2', '3.3'] 
 cpuCores = '4' 
 cpuFlags = 
 'fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,dts,acpi,mmx,fxsr,sse,sse2,ss,ht,tm,pbe,syscall,nx,pdpe1gb,rdtscp,lm,constant_tsc,arch_perfmon,pebs,bts,rep_good,nopl,xtopology,nonstop_tsc,aperfmperf,eagerfpu,pni,pclmulqdq,dtes64,monitor,ds_cpl,vmx,smx,est,tm2,ssse3,cx16,xtpr,pdcm,pcid,dca,sse4_1,sse4_2,x2apic,popcnt,tsc_deadline_timer,aes,xsave,avx,lahf_lm,ida,arat,epb,xsaveopt,pln,pts,dtherm,tpr_shadow,vnmi,flexpriority,ept,vpid,model_Nehalem,model_Conroe,model_coreduo,model_core2duo,model_Penryn,model_Westmere,model_n270,model_SandyBridge'
  
 cpuModel = 'Intel(R) Xeon(R) CPU E5-1620 0 @ 3.60GHz' 
 cpuSockets = '1' 
 cpuSpeed = '3708.000' 
 cpuThreads = '8' 
 emulatedMachines = ['pc', 
 'q35', 
 'isapc', 
 'pc-0.10', 
 'pc-0.11', 
 'pc-0.12', 
 'pc-0.13', 
 'pc-0.14', 
 'pc-0.15', 
 'pc-1.0', 
 'pc-1.1', 
 'pc-1.2', 
 'pc-1.3', 
 'none'] 
 guestOverhead = '65' 
 hooks = {} 
 kvmEnabled = 'true' 
 lastClient = '192.168.1.41' 
 lastClientIface = 'em1' 
 management_ip = '0.0.0.0' 
 memSize = '16001' 
 netConfigDirty = 'True' 
 networks = {} 
 nics = {'em1': {'addr': '192.168.1.42', 
 'cfg': {}, 
 'hwaddr': '00:1a:6b:51:de:b4', 
 'ipv6addrs': ['fe80::21a:6bff:fe51:deb4/64'], 
 'mtu': '1500', 
 'netmask': '255.255.255.0', 
 'speed': 100}} 
 operatingSystem = {'name': 'Fedora', 'release': '4', 'version': '19'} 
 packages2 = {'kernel': {'buildtime': 1384978944.0, 
 'release': '200.fc19.x86_64', 
 'version': '3.11.9'}, 
 'libvirt': {'buildtime': 1384730741, 
 'release': '2.fc19', 
 'version': '1.0.5.7'}, 
 'mom': {'buildtime': 1375215820, 'release': '3.fc19', 'version': '0.3.2'}, 
 'qemu-img': {'buildtime': 1383700301, 
 'release': '13.fc19', 
 'version': '1.4.2'}, 
 'qemu-kvm': {'buildtime': 1383700301, 
 'release': '13.fc19', 
 'version': '1.4.2'}, 
 'spice-server': {'buildtime': 1383130020, 
 'release': '3.fc19', 
 'version': '0.12.4'}, 
 'vdsm': {'buildtime': 1384274283, 'release': '11.fc19', 'version': '4.13.0'}} 
 reservedMem = '321' 
 software_revision = '11' 
 software_version = '4.13' 
 supportedENGINEs = ['3.0', '3.1', '3.2', '3.3'] 
 supportedProtocols = ['2.2', '2.3'] 
 uuid = '0A583269-811F-E211-AA06-001A6B51DEB4' 
 version_name = 'Snow Man' 
 vlans = {} 
 vmTypes = ['kvm'] 
 [ root@lab2 vdsm]# 
 
 
 Pascal Jakobi 
 116 rue de Stalingrad 
 93100 Montreuil, France 
 Tel : + 33 6 87 47 58 19 
 @ : pjak...@yahoo.fr 
 
 
 Le Lundi 2 décembre 2013 10h31, Pascal Jakobi  pascal.jak...@gmail.com  a 
 écrit : 
 As a matter of fact, when I installed F19, it created an eno1 device, aliased 
 to em1. Don't know why, yet. 
 Even adding a GATEWAY clause in /etc/sysconfig/network does

Re: [Users] Keepalived on oVirt Hosts has engine networking issues

2013-12-01 Thread Assaf Muller
Could you please attach the output of:
vdsClient -s 0 getVdsCaps
(Or without the -s, whichever works)
And:
ip a

On both hosts?
You seem to have made changes since the documentation on the link you provided, 
like separating the management and storage via VLANs on eth0. Any other changes?


Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Andrew Lau and...@andrewklau.com
To: users users@ovirt.org
Sent: Sunday, December 1, 2013 4:55:32 AM
Subject: [Users] Keepalived on oVirt Hosts has engine networking issues

Hi, 

I have the scenario where I have gluster and ovirt hosts on the same box, to 
keep the gluster volumes highly available incase a box drops I'm using 
keepalived across the boxes and using that IP as the means for the storage 
domain. I documented my setup here in case anyone needs a little more info 
http://www.andrewklau.com/returning-to-glusterized-ovirt-3-3/ 

However, the engine seems to be picking up the floating IP assigned to 
keepalived as the interface and messing with the ovirtmgmt migration network, 
so migrations are failing as my floating IP gets assigned to the ovirtmgmt 
bridge in the engine however it's not actually there on most hosts (except one) 
so vdsm seems to report destination same as source. 

I've since created a new vlan interface just for storage to avoid the ovirtmgmt 
conflict, but the engine will still pick up the wrong IP on the storage vlan 
because of keepalived. This means I can't use the save network feature within 
the engine as it'll save the floating ip rather than the one already there. Is 
this a bug or just the way it's designed. 

eth0.2 - ovirtmgmt (172.16.0.11) - management and migration network - engine 
sees, sets and saves 172.16.0.11 
eth0.3 - storagenetwork (172.16.1.11) - gluster network - engine sees, sets 
and saves 172.16.1.5 (my floating IP) 

I hope this makes sense. 

p.s. can anyone also confirm, does gluster support multi pathing by default? If 
I'm using this keepalived method, am I bottle necking myself to one host? 

Thanks, 
Andrew 

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] ovirtmgmt not installed

2013-12-01 Thread Assaf Muller
Did you install VDSM from nightly or stable?

Assaf Muller, Cloud Networking Engineer 
Red Hat 

- Original Message -
From: Pascal Jakobi pascal.jak...@gmail.com
To: Mike Kolesnik mkole...@redhat.com
Cc: users@ovirt.org
Sent: Sunday, December 1, 2013 10:58:40 AM
Subject: Re: [Users] ovirtmgmt not installed

Mike 

Here you go. However, please note that I must investigate the connection issue 
that Alon saw. will do it tomorrow. 
Many thanks to you folks. 
P 

[root@lab2 ~]# vdsClient -s 0 getVdsCaps 
HBAInventory = {'FC': [], 'iSCSI': [{'InitiatorName': 
'iqn.1994-05.com.redhat:eea139a8'}]} 
ISCSIInitiatorName = 'iqn.1994-05.com.redhat:eea139a8' 
bondings = {'bond0': {'addr': '', 
'cfg': {}, 
'hwaddr': 'fa:7e:79:56:5a:c2', 
'ipv6addrs': [], 
'mtu': '1500', 
'netmask': '', 
'slaves': []}} 
bridges = {} 
clusterLevels = ['3.0', '3.1', '3.2', '3.3'] 
cpuCores = '4' 
cpuFlags = 
'fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,dts,acpi,mmx,fxsr,sse,sse2,ss,ht,tm,pbe,syscall,nx,pdpe1gb,rdtscp,lm,constant_tsc,arch_perfmon,pebs,bts,rep_good,nopl,xtopology,nonstop_tsc,aperfmperf,eagerfpu,pni,pclmulqdq,dtes64,monitor,ds_cpl,vmx,smx,est,tm2,ssse3,cx16,xtpr,pdcm,pcid,dca,sse4_1,sse4_2,x2apic,popcnt,tsc_deadline_timer,aes,xsave,avx,lahf_lm,ida,arat,epb,xsaveopt,pln,pts,dtherm,tpr_shadow,vnmi,flexpriority,ept,vpid,model_Nehalem,model_Conroe,model_coreduo,model_core2duo,model_Penryn,model_Westmere,model_n270,model_SandyBridge'
 
cpuModel = 'Intel(R) Xeon(R) CPU E5-1620 0 @ 3.60GHz' 
cpuSockets = '1' 
cpuSpeed = '3744.000' 
cpuThreads = '8' 
emulatedMachines = ['pc', 
'q35', 
'isapc', 
'pc-0.10', 
'pc-0.11', 
'pc-0.12', 
'pc-0.13', 
'pc-0.14', 
'pc-0.15', 
'pc-1.0', 
'pc-1.1', 
'pc-1.2', 
'pc-1.3', 
'none'] 
guestOverhead = '65' 
hooks = {} 
kvmEnabled = 'false' 
lastClient = '192.168.1.41' 
lastClientIface = 'em1' 
management_ip = '0.0.0.0' 
memSize = '16001' 
netConfigDirty = 'False' 
networks = {} 
nics = {'em1': {'addr': '192.168.1.42', 
'cfg': {}, 
'hwaddr': '00:1a:6b:51:de:b4', 
'ipv6addrs': ['fe80::21a:6bff:fe51:deb4/64'], 
'mtu': '1500', 
'netmask': '255.255.255.0', 
'speed': 100}} 
operatingSystem = {'name': 'Fedora', 'release': '2', 'version': '19'} 
packages2 = {'kernel': {'buildtime': 1384978944.0, 
'release': '200.fc19.x86_64', 
'version': '3.11.9'}, 
'libvirt': {'buildtime': 1384730741, 
'release': '2.fc19', 
'version': '1.0.5.7'}, 
'mom': {'buildtime': 1375215820, 'release': '3.fc19', 'version': '0.3.2'}, 
'qemu-img': {'buildtime': 1383700301, 
'release': '13.fc19', 
'version': '1.4.2'}, 
'qemu-kvm': {'buildtime': 1383700301, 
'release': '13.fc19', 
'version': '1.4.2'}, 
'spice-server': {'buildtime': 1383130020, 
'release': '3.fc19', 
'version': '0.12.4'}, 
'vdsm': {'buildtime': 1384274283, 'release': '11.fc19', 'version': '4.13.0'}} 
reservedMem = '321' 
software_revision = '11' 
software_version = '4.13' 
supportedENGINEs = ['3.0', '3.1', '3.2', '3.3'] 
supportedProtocols = ['2.2', '2.3'] 
uuid = '0A583269-811F-E211-AA06-001A6B51DEB4' 
version_name = 'Snow Man' 
vlans = {} 
vmTypes = ['kvm'] 



2013/12/1 Mike Kolesnik  mkole...@redhat.com  







Hi there 

Hi Pascal, 






I installed a console on F19, then a F19 host (time 11:09 today). 
Everything works fine, apart from the installation of the mgmt network at the 
end. 
Can someone tell me what's going wrong ? 

Can you please send the output of vdsCaps from the host (vdsClient -s 0 
getVdsCaps)? 




Thxs in advance 
Pascal 




-- 
Pascal Jakobi 
116 rue de Stalingrad 
93100 Montreuil, France 
+33 6 87 47 58 19 
pascal.jak...@gmail.com 

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Failed to configure management network on the host

2013-11-13 Thread Assaf Muller
Please attach vdsm.log and supervdsm.log after you get the Failed to configure 
management network on the host error.

- Original Message -
From: Rob Abshear rabsh...@citytwist.net
To: users@ovirt.org
Sent: Wednesday, November 13, 2013 6:12:57 PM
Subject: [Users] Failed to configure management network on the host

I am running oVirt Engine Version: 
3.4.0-0.2.master.20131107144332.git214186b.fc18.  I am having an issue 
adding new hosts.  I do a fresh install of fc18 on the host and then try 
to add it into my cluster.  The controller installs all packages and it 
looks like it's all good, but at the end I get the message Failed to 
configure management network on the host and the host goes into non 
operational state.  I can remove the host and add it into a second oVirt 
environment that I am running.  This other environment is runningoVirt 
Engine Version: 3.2.3-1.fc18.  Once I add it into that other 
environment, I can then remove it and successfully add it into my 3.4 
environment.

-- 


*NADA Convention  Expo*:  January 24-27, 2014 - New Orleans, LA - Booth 
5001
http://www.bbb.org/south-east-florida/business-reviews/advertising-agencies/citytwist-in-boca-raton-fl-90033448http://www.bbb.org/south-east-florida/business-reviews/advertising-agencies/citytwist-in-boca-raton-fl-90033448
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


[Users] Default route on hosts

2013-11-12 Thread Assaf Muller
Hi Christopher,

Your understanding of the terms is correct. 
A small clarification - Hosts are mostly referred to as nodes when running 
oVirt Node, a custom Linux distribution purposely made to be used as a host in 
oVirt. 

As for your question: 
DEFROUTE=yes/no in the hosts' ifcfg files control the default gateway of the 
host. Currently, by design, the management network's gateway (Which sounds like 
your management network doesn't have?) will be kept as the host's default 
gateway. This was done to make sure that the engine maintains connectivity to 
its hosts. 

oVirt 3.3 introduced a feature called multiple gateways which may assist you. 
If you configured a gateway on the public network (Either statically via oVirt 
or via DHCP), then all traffic into the hosts through their public network NIC 
will be returned via that nic, *even though* the host's default gateway is 
ovirtmgmt's gateway. 

To clarify further: The current behavior is that the host's default gateway is 
ovirtmgmt's gateway. This controls traffic originating from *within* the host 
such as communications with storage. Incoming traffic, however, such as traffic 
originating from users from outside of the datacenter, will be returned via 
the device that the traffic came in on. 

If that doesn't work for you, and you feel like you need to control the host's 
default gateway via oVirt, then please file an RFE. I can't guarantee a 
timeline 
or that it will ultimately make it in the product, but the RFE will be 
carefully taken into account when designing future versions.

- Original Message -

From: Christopher Geddings chris.geddi...@duke.edu 
To: users@ovirt.org 
Sent: Monday, November 11, 2013 8:40:33 PM 
Subject: [Users] Default route on hosts 

Before I ask my question, I want to make sure I'm using the right terms: 
engine == the system that manages the ovirt environment. Runs the web 
interface, etc 
host(s) == the physical machine(s) the engine uses to deploy virtual machines 
on. Sometimes referenced as a node? 
vm(s) == the guest systems that are controlled by the engine and run on the 
hosts, that do the stuff you present to the world at large, whatever your world 
at large would be. 

Is that correct? 

In my environment, I have a customer facing network and another network that's 
on a set of isolated switches. I desire to use the isolated 
network as the management network, and am naming the other network public in 
the web management interface. 

In the interface, I see these labeled as ovirtmgmt and public, for the 
management network (private) and the customer facing network (public). 

From what I can tell from poking at the configuration files, and from a couple 
references in other threads, the 'DEFROUTE' setting in the ifcfg-* files 
gets set to 'yes' only for the ovirtmgmt network, not for any other bridged 
network. 

I can override this by hand as I deploy hosts, but am hoping for a way to do so 
within the ovirt management framework, so I have less bits that can 
go wrong. Is there a way for me to get the DEFROUTE set to no for the ovirtmgmt 
interface and to yes on the public interface as I deploy hosts? 

Is there a reason I'm not seeing for avoiding this behavior? 

--Chris 


___ 
Users mailing list 
Users@ovirt.org 
http://lists.ovirt.org/mailman/listinfo/users 
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Low quality of el6 vdsm rpms

2013-11-12 Thread Assaf Muller
Regarding the pep8 breakage - Try updating your pep8.

- Original Message -
From: Patrick Hurrelmann patrick.hurrelm...@lobster.de
To: oVirt Mailing List users@ovirt.org
Sent: Tuesday, November 12, 2013 11:34:20 AM
Subject: [Users] Low quality of el6 vdsm rpms

Hi all,

sorry for this rant, but...

I now tried several times to test the beta 3.3.1 rpms, but they can't
even be installed in the most times. One time it required a future
selinux-policy, although the needed selinux fix was delivered in a much
lower version. Now the rpms have broken requirements. It requires
hostname instead of /bin/hostname. This broken requirement is not
included in the vdsm 3.3 branch, so I wonder where it comes from?
Anyway. So I proceeded and tried to build vdsm myself once again.
Currently the build fails with (but worked fine some days ago):

/usr/bin/pep8 --exclude=config.py,constants.py --filename '*.py,*.py.in' \
client lib/cpopen/*.py lib/vdsm/*.py lib/vdsm/*.py.in tests
vds_bootstrap vdsm-tool vdsm/*.py vdsm/*.py.in vdsm/netconf
vdsm/sos/vdsm.py.in vdsm/storage vdsm/vdsm vdsm_api vdsm_hooks vdsm_reg
vdsm/storage/imageRepository/formatConverter.py:280:29: E128
continuation line under-indented for visual indent


- How can the quality of the vdsm builds be increased? It is frustrating
to spend time on testing and then the hosts cannot even be installed to
broken vdsm rpms.
- How are the builds prepared? Is there a Jenkins job that prepares
stable rpms in addition to the nightly job? Or is this totally
handcrafted?
- How can it be that the rpm spec differs between the 3.3 branch and
released rpms? What is the source/branch for el6 vdsm rpms? Maybe I'm
just tracking on the wrong source tree...

Thx and Regards
Patrick


-- 
Lobster LOGsuite GmbH, Münchner Straße 15a, D-82319 Starnberg

HRB 178831, Amtsgericht München
Geschäftsführer: Dr. Martin Fischer, Rolf Henrich
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Default route on hosts

2013-11-12 Thread Assaf Muller
Can users outside of the hosts' networks reach the VMs in the hosts?

If you use netstat -rn it is expected that the gateway will be 0.0.0.0, as 
ifcfg-ovirtmgmt has DEFROUTE=yes and ifcfg-public has DEFROUTE=no, then 
ovirtmgmt's
'gateway' (0.0.0.0) will be determined as the host's default gateway. However 
with the new multiple gateways feature we configure source routing to make
sure that traffic that comes (from the outside) in the public network's device 
will return the way it came in.

You can use 'ip rule' to see the rules VDSM configures. It creates two rules 
and a routing table per device. You can use 'ip route show table %s' on each
table, where the IDs can be obtained by 'ip rule'.

- Original Message -
From: Chris Geddings chris.geddi...@duke.edu
To: Assaf Muller amul...@redhat.com
Sent: Tuesday, November 12, 2013 2:26:40 PM
Subject: Re: [Users] Default route on hosts


On Nov 12, 2013, at 3:32 AM, Assaf Muller amul...@redhat.com wrote:

 oVirt 3.3 introduced a feature called multiple gateways which may assist you. 
 If you configured a gateway on the public network (Either statically via oVirt
 or via DHCP), then all traffic into the hosts through their public network 
 NIC will be returned via that nic, *even though* the host's default gateway is
 ovirtmgmt's gateway.
 
So, interestingly or not, when I define a gateway on the public interface 
(through the web management interface or DHCP), and I don't worry about making 
the
DEFROUTE=yes setting in my public network, the box behaves like it has an 
incorrect default route.  Now, my management network has a gateway
of 0.0.0.0, as it is a completely simple network, so that may be part of the 
problem. I'm not sure the impact of a 0.0.0.0 as a gateway.

I'm not sure how to poke at this further to figure out where the breakage is.  
Routing looks like it has gotten a little more complex, and I'm still operating 
with 
'netstat -rn' and 'route add foo' type commands.

--Chris

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] Tagged and Untagged Traffic on the same Interface.

2013-11-03 Thread Assaf Muller
Hi Matt,

You can associate multiple networks with a single nic / bond, providing that 
the untagged network is also non-VM.
You may have up to one untagged non-VM network, and zero or more tagged 
networks on the same network device.

Example - All networks on the same device:
Non-VM, untagged management network
Tagged, non-VM network X 3
Tagged, VM network X 2

- Original Message -
From: Matt Curry mcu...@skopos.us
To: users@ovirt.org
Sent: Saturday, November 2, 2013 12:11:04 AM
Subject: [Users] Tagged and Untagged Traffic on the same Interface.

Hello All, 

Is it possible to have tagged, and untagged traffic on the same interface. 

Example. 
2 nics on node. 
Bond them together. 

Management traffic and other traffic on same bonded interface? 

All help is appreciated. 

PS. I am on #ovirt as MCLinux… 
Thanks 


This is a PRIVATE message. If you are not the intended recipient, please delete 
without copying and kindly advise us by e-mail of the mistake in delivery. 
NOTE: Regardless of content, this e-mail shall not operate to bind SKOPOS to 
any order or other contract unless pursuant to explicit written agreement or 
government initiative expressly permitting the use of e-mail for such purpose. 

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] received packet with own address as source address

2013-08-22 Thread Assaf Muller
I'd enable STP on all ports, including the port connecting your switch to the 
school switch.

VSDSM configures the ovirtmgmt bridge on installation. The reason I'm 
suggesting that it's not VDSM at fault but rather the school network having a 
loop is that you mentioned
that you get these messages even with no VMs running.

- Original Message -
From: H. Haven Liu haven@ucla.edu
To: Assaf Muller amul...@redhat.com
Cc: Users@ovirt.org
Sent: Wednesday, August 21, 2013 11:26:57 PM
Subject: Re: [Users] received packet with own address as source address

The switch is connected to the internet via the school's switch. Currently (as 
well as when I received the error messages last time) portfast is disabled for 
all ports, and STP enabled for all ports except for the port connecting my 
switch to the school's.

On Aug 20, 2013, at 11:50 PM, Assaf Muller amul...@redhat.com wrote:

 The problem was also present when I just had one interface connected to the 
 switch, connected to the Internet.
 
 How was the switch connected to the internet? Is it connected to other 
 switches?
 
 To solve the loop you can make sure that STP is enabled on all switches and 
 to disable portfast.
 
 - Original Message -
 From: H. Haven Liu haven@ucla.edu
 To: Assaf Muller amul...@redhat.com
 Cc: Users@ovirt.org
 Sent: Tuesday, August 20, 2013 6:09:09 PM
 Subject: Re: [Users] received packet with own address as source address
 
 Hello,
 
 After I reinstalled the host, the problem seems to have disappeared. This is 
 not to say it's not going to come back, as it has done before. 
 
 I know the nodes were not receiving the error message *before* I installed 
 VDSM (which configured the bridge?), but I haven't tried uninstalling to see 
 if the problem would go away. The problem was also present when I just had 
 one interface connected to the switch, connected to the Internet.
 
 If the problem is indeed a classic L2 loop, how do I go about resolving it? 
 Is it due to bridge (mis)configuration, or is it in the network switch 
 settings? I have tried enabling and disabling STP and portfast, which didn't 
 seem to have any effect.
 
 Thanks,
 
 Haven
 On Aug 20, 2013, at 12:26 AM, Assaf Muller amul...@redhat.com wrote:
 
 Hello Mr. Liu,
 
 Sounds like a classic L2 loop. Hosts getting their own packets then losing 
 connectivity for a few minutes while the switches restart.
 
 Can you try to rule out this possibility?
 For example:
 If you uninstall the engine and VDSM from all hosts does this still happen?
 Can you simplify the network topology and see if the issue persists?
 
 
 - Original Message -
 From: H. Haven Liu haven@ucla.edu
 To: Users@ovirt.org
 Sent: Saturday, August 17, 2013 10:06:07 PM
 Subject: [Users] received packet with own address as source address
 
 Hello,
 
 I have a oVirt installation with 6 CentOS 6.4 hosts. Some (not all) hosts 
 are receiving these error messages, at a frequency of once every other 
 minute:
 
 [root@vhost1 ~]# grep source address /var/log/messages | tail
 Aug 17 11:19:20 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:21:24 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:23:29 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:25:34 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:27:39 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:29:44 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:31:49 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:33:54 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:35:59 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:38:04 vhost1 kernel: em1: received packet with own address as 
 source address
 
 These error messages are accompanied with loss of connectivity, with 
 duration varies between a few seconds to minutes. The loss of connectivity 
 would start just before the above error message.
 
 I don't *think* it's hardware because the received packet with own address 
 error message would disappear from one of the hosts after a while. These 
 error messages are also present whether or not any VM is running on the 
 particular host.
 
 Here are some configurations:
 
 [root@vhost1 ~]# ifconfig
 em1   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f1  
 inet6 addr: :::::eee1/64 Scope:Link
 UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
 RX packets:37413202 errors:0 dropped:0 overruns:0 frame:0
 TX packets:17822315 errors:0 dropped:0 overruns:0 carrier:0
 collisions:0 txqueuelen:1000 
 RX bytes:32444347931 (30.2 GiB)  TX bytes:13715272213 (12.7 GiB)
 Interrupt:16 
 
 em2   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f2  
 inet6 addr

Re: [Users] received packet with own address as source address

2013-08-21 Thread Assaf Muller
 The problem was also present when I just had one interface connected to the 
 switch, connected to the Internet.

How was the switch connected to the internet? Is it connected to other switches?

To solve the loop you can make sure that STP is enabled on all switches and to 
disable portfast.

- Original Message -
From: H. Haven Liu haven@ucla.edu
To: Assaf Muller amul...@redhat.com
Cc: Users@ovirt.org
Sent: Tuesday, August 20, 2013 6:09:09 PM
Subject: Re: [Users] received packet with own address as source address

Hello,

After I reinstalled the host, the problem seems to have disappeared. This is 
not to say it's not going to come back, as it has done before. 

I know the nodes were not receiving the error message *before* I installed VDSM 
(which configured the bridge?), but I haven't tried uninstalling to see if the 
problem would go away. The problem was also present when I just had one 
interface connected to the switch, connected to the Internet.

If the problem is indeed a classic L2 loop, how do I go about resolving it? Is 
it due to bridge (mis)configuration, or is it in the network switch settings? I 
have tried enabling and disabling STP and portfast, which didn't seem to have 
any effect.

Thanks,

Haven
On Aug 20, 2013, at 12:26 AM, Assaf Muller amul...@redhat.com wrote:

 Hello Mr. Liu,
 
 Sounds like a classic L2 loop. Hosts getting their own packets then losing 
 connectivity for a few minutes while the switches restart.
 
 Can you try to rule out this possibility?
 For example:
  If you uninstall the engine and VDSM from all hosts does this still happen?
  Can you simplify the network topology and see if the issue persists?
 
 
 - Original Message -
 From: H. Haven Liu haven@ucla.edu
 To: Users@ovirt.org
 Sent: Saturday, August 17, 2013 10:06:07 PM
 Subject: [Users] received packet with own address as source address
 
 Hello,
 
 I have a oVirt installation with 6 CentOS 6.4 hosts. Some (not all) hosts are 
 receiving these error messages, at a frequency of once every other minute:
 
 [root@vhost1 ~]# grep source address /var/log/messages | tail
 Aug 17 11:19:20 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:21:24 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:23:29 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:25:34 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:27:39 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:29:44 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:31:49 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:33:54 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:35:59 vhost1 kernel: em1: received packet with own address as 
 source address
 Aug 17 11:38:04 vhost1 kernel: em1: received packet with own address as 
 source address
 
 These error messages are accompanied with loss of connectivity, with duration 
 varies between a few seconds to minutes. The loss of connectivity would start 
 just before the above error message.
 
 I don't *think* it's hardware because the received packet with own address 
 error message would disappear from one of the hosts after a while. These 
 error messages are also present whether or not any VM is running on the 
 particular host.
 
 Here are some configurations:
 
 [root@vhost1 ~]# ifconfig
 em1   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f1  
  inet6 addr: :::::eee1/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:37413202 errors:0 dropped:0 overruns:0 frame:0
  TX packets:17822315 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:1000 
  RX bytes:32444347931 (30.2 GiB)  TX bytes:13715272213 (12.7 GiB)
  Interrupt:16 
 
 em2   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f2  
  inet6 addr: :::::eee2/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:21878 errors:0 dropped:0 overruns:0 frame:0
  TX packets:526 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:1000 
  RX bytes:1598070 (1.5 MiB)  TX bytes:40526 (39.5 KiB)
  Interrupt:17 
 
 loLink encap:Local Loopback  
  inet addr:127.0.0.1  Mask:255.0.0.0
  inet6 addr: ::1/128 Scope:Host
  UP LOOPBACK RUNNING  MTU:16436  Metric:1
  RX packets:49 errors:0 dropped:0 overruns:0 frame:0
  TX packets:49 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:0 
  RX bytes:9236 (9.0 KiB)  TX bytes:9236 (9.0 KiB)
 
 ovirtmgmt Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f1  
  inet addr:xxx.xxx.xxx.xxx  Bcast:xxx.xxx.xxx.yyy  Mask:255.255.255.0
  inet6 addr

Re: [Users] received packet with own address as source address

2013-08-20 Thread Assaf Muller
Hello Mr. Liu,

Sounds like a classic L2 loop. Hosts getting their own packets then losing 
connectivity for a few minutes while the switches restart.

Can you try to rule out this possibility?
For example:
  If you uninstall the engine and VDSM from all hosts does this still happen?
  Can you simplify the network topology and see if the issue persists?


- Original Message -
From: H. Haven Liu haven@ucla.edu
To: Users@ovirt.org
Sent: Saturday, August 17, 2013 10:06:07 PM
Subject: [Users] received packet with own address as source address

Hello,

I have a oVirt installation with 6 CentOS 6.4 hosts. Some (not all) hosts are 
receiving these error messages, at a frequency of once every other minute:

[root@vhost1 ~]# grep source address /var/log/messages | tail
Aug 17 11:19:20 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:21:24 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:23:29 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:25:34 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:27:39 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:29:44 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:31:49 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:33:54 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:35:59 vhost1 kernel: em1: received packet with own address as source 
address
Aug 17 11:38:04 vhost1 kernel: em1: received packet with own address as source 
address

These error messages are accompanied with loss of connectivity, with duration 
varies between a few seconds to minutes. The loss of connectivity would start 
just before the above error message.

I don't *think* it's hardware because the received packet with own address 
error message would disappear from one of the hosts after a while. These error 
messages are also present whether or not any VM is running on the particular 
host.

Here are some configurations:

[root@vhost1 ~]# ifconfig
em1   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f1  
  inet6 addr: :::::eee1/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:37413202 errors:0 dropped:0 overruns:0 frame:0
  TX packets:17822315 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:1000 
  RX bytes:32444347931 (30.2 GiB)  TX bytes:13715272213 (12.7 GiB)
  Interrupt:16 

em2   Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f2  
  inet6 addr: :::::eee2/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:21878 errors:0 dropped:0 overruns:0 frame:0
  TX packets:526 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:1000 
  RX bytes:1598070 (1.5 MiB)  TX bytes:40526 (39.5 KiB)
  Interrupt:17 

loLink encap:Local Loopback  
  inet addr:127.0.0.1  Mask:255.0.0.0
  inet6 addr: ::1/128 Scope:Host
  UP LOOPBACK RUNNING  MTU:16436  Metric:1
  RX packets:49 errors:0 dropped:0 overruns:0 frame:0
  TX packets:49 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:0 
  RX bytes:9236 (9.0 KiB)  TX bytes:9236 (9.0 KiB)

ovirtmgmt Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f1  
  inet addr:xxx.xxx.xxx.xxx  Bcast:xxx.xxx.xxx.yyy  Mask:255.255.255.0
  inet6 addr: :::::/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:22623881 errors:0 dropped:0 overruns:0 frame:0
  TX packets:9990988 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:0 
  RX bytes:28427312067 (26.4 GiB)  TX bytes:13112553784 (12.2 GiB)

private_lan Link encap:Ethernet  HWaddr aa:bb:cc:dd:ee:f2  
  inet addr:10.0.1.12  Bcast:10.0.1.255  Mask:255.255.255.0
  inet6 addr: :::::eee2/64 Scope:Link
  UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
  RX packets:19771 errors:0 dropped:0 overruns:0 frame:0
  TX packets:6 errors:0 dropped:0 overruns:0 carrier:0
  collisions:0 txqueuelen:0 
  RX bytes:1080512 (1.0 MiB)  TX bytes:468 (468.0 b)
[root@vhost1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-em1
DEVICE=em1
ONBOOT=yes
HWADDR=aa:bb:cc:dd:ee:f2
BRIDGE=ovirtmgmt
NM_CONTROLLED=no
UUID=c25b1ad1-9646-4126-9149-be5f68bae36c
DNS2=www.www.www.www
DNS1=www.www.www.www
HOSTNAME=vhost1.mydomain.com
MTU=1500
IPV6INIT=no
[root@vhost1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-ovirtmgmt 
DEVICE=ovirtmgmt
ONBOOT=yes
TYPE=Bridge
DELAY=0
IPADDR=xxx.xxx.xxx.xxx
NETMASK=255.255.255.0
GATEWAY=xxx.xxx.xxx.1
BOOTPROTO=none
NM_CONTROLLED=no
UUID=c25b1ad1-9646-4126-9149-be5f68bae36c
DNS2=www.www.www.www

Re: [Users] How to add internal networks

2013-05-13 Thread Assaf Muller
Hi Frank,

Without the ability to control the networking devices in your vicinity and 
setup VLANs, what you're trying to do isn't supported via oVirt.

However, assuming you do have root access to your host, I can suggest a 
workaround:
* Create two dummy interfaces on the host, and give them static IP addresses. 
For example: 192.168.1.254/24, and 192.168.2.254/24.
* Create two networks (Non-VM) called One and Two and attach them to the dummy 
NICs.
* Create VMs with virtual NICs in network One, and VMs with virtual NICs in 
network Two. 
  You can give them static IP addresses and supply the host's dummy interfaces 
as gateways, or enable a DHCP server on the host.
* Enable routing on the host, and setup the appropriate rules so that he acts 
as a default gateway for all of the VMs, and that he forwards traffic 
appropriately between the VM networks.

- Original Message -
From: Frank Wall f...@moov.de
To: users@ovirt.org
Sent: Monday, May 13, 2013 2:06:53 PM
Subject: Re: [Users] How to add internal networks

Hi René,

Am 13.05.2013 09:56, schrieb René Koch (ovido):
 When using multiple logical networks on 1 physical nic you
 will need most likely VLAN tagging, as all vm networks are
 configured as a bridge in oVirt.

unfortunately, using VLAN tagging is not an option for me,
because I'm not the maintainer of the network devices (and
my hosting provider won't configure VLANs for me).

Is there another way to get multiple networks in oVirt up
and running?

FYI: The main purpose of these internal networks is a test
environment for some VMs. Those VMs should be able to reach
each other and to have internet access.


Thanks
- Frank
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [Users] How to add internal networks

2013-05-13 Thread Assaf Muller
Correction:
The two networks should be VM networks.

- Original Message -
From: Assaf Muller amul...@redhat.com
To: Frank Wall f...@moov.de
Cc: users@ovirt.org
Sent: Monday, May 13, 2013 4:05:16 PM
Subject: Re: [Users] How to add internal networks

Hi Frank,

Without the ability to control the networking devices in your vicinity and 
setup VLANs, what you're trying to do isn't supported via oVirt.

However, assuming you do have root access to your host, I can suggest a 
workaround:
* Create two dummy interfaces on the host, and give them static IP addresses. 
For example: 192.168.1.254/24, and 192.168.2.254/24.
* Create two networks (Non-VM) called One and Two and attach them to the dummy 
NICs.
* Create VMs with virtual NICs in network One, and VMs with virtual NICs in 
network Two. 
  You can give them static IP addresses and supply the host's dummy interfaces 
as gateways, or enable a DHCP server on the host.
* Enable routing on the host, and setup the appropriate rules so that he acts 
as a default gateway for all of the VMs, and that he forwards traffic 
appropriately between the VM networks.

- Original Message -
From: Frank Wall f...@moov.de
To: users@ovirt.org
Sent: Monday, May 13, 2013 2:06:53 PM
Subject: Re: [Users] How to add internal networks

Hi René,

Am 13.05.2013 09:56, schrieb René Koch (ovido):
 When using multiple logical networks on 1 physical nic you
 will need most likely VLAN tagging, as all vm networks are
 configured as a bridge in oVirt.

unfortunately, using VLAN tagging is not an option for me,
because I'm not the maintainer of the network devices (and
my hosting provider won't configure VLANs for me).

Is there another way to get multiple networks in oVirt up
and running?

FYI: The main purpose of these internal networks is a test
environment for some VMs. Those VMs should be able to reach
each other and to have internet access.


Thanks
- Frank
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users