- Users - oVirt List Archives

Spice on Ovirt 3.4.4 Windows 2012 Guests
by Gary Lloyd 16 Oct '14

16 Oct '14

Hi We are currently running Ovirt 3.3.5 in production, currently our Windows Guests are using spice for the console. I have noticed that in our test environment running 3.4.4 that this option cannot be selected for Windows 2012 guests. Does anyone know what will happen to the console on VMs running windows 2012 if we upgrade to 3.4.4 on our production environment ? Will it simply not work resulting in us having to power down the affected VMs or will the existing VMs carry on working with the spice console setting ? Thanks *Gary Lloyd* ---------------------------------- IT Services Keele University -----------------------------------

2 2

qemu-rhev in centos 7
by Nathanaël Blanchet 15 Oct '14

15 Oct '14

Hello, native el7 qemu-kvm package is 1.2.x so that it is supposed to contain the missing flag in el6 0.12 one for live migration. I saw that qemu-kvm-rhev.el7 was also provided in the ovirt repo, but may be unuseful, in this case... what do you think ?(I didn't try livre migration yet with el7 native qemu)

2 2

delete snapshots or create a template make a huge impact on VMs on SPM-node
by Ricky Schneberger 15 Oct '14

15 Oct '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, When we do some tasks in ovirt as delete an snapshot or clone a machine or just want to make a machine template the impact on VMs residing on the SPM node is huge. Is there a way to prevent this? This is just normal tasks and we do this all the time. - -- Ricky Schneberger - ------------------------------------ DANGER! Human at keyboard! -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlQkA0AACgkQOap81biMC2MfdwCdEe4PJs0OdQwqsrNVg4qmGwuJ Tp4An0rnDMjHf2JcS/lC7d4pBb3G9Jzw =5yO2 -----END PGP SIGNATURE-----

4 3

Re: [ovirt-users] ovirt network VLAN tagging
by Juan Pablo Lorier 15 Oct '14

15 Oct '14

Hi, It seems to me that it may be a switch configuration problem. The vms in the same host can talk to each other via the bridge, but if your switch is not configured to allow tagged traffic for vlans 50 and 60 (and all the way to the interface of the other host must allow tagged traffic for those vlans), you won't be able to get traffic. Regards,

1 0

Space not getting freed in storage domain after removing vm
by Chandrahasa S 15 Oct '14

15 Oct '14

Dear All, I have only one VM in my Storage domain of size actual is 1 GB, but my Storage domain show free 288 GB out of 499 GB. Any idea what is the issue. Thanks & Regards Chandrahasa S =====-----=====-----===== Notice: The information contained in this e-mail message and/or attachments to it may contain confidential or privileged information. If you are not the intended recipient, any dissemination, use, review, distribution, printing or copying of the information contained in this e-mail message and/or attachments to it are strictly prohibited. If you have received this communication in error, please notify us by reply e-mail or telephone and immediately and permanently delete the message and any attachments. Thank you

2 1

async vm start (python)
by Giulio Casella 15 Oct '14

15 Oct '14

Hi everybody, I'm trying to start a VM (using python sdk) asynchronously, but I obtain synchronous behaviour. The engine is RHEV-M version 3.4.1-0.31.el6ev. Here a snippet of code: --------------------------------------------------------- myvm = api.vms.get(name="VMNAME") retval = myvm.start(params.Action(async=True)) print ParseHelper.toXml(retval) --------------------------------------------------------- and this is what I get (retval): --------------------------------------------------------- <action> <async>true</async> <vm href="/api/vms/5aa66e96-5fdd-4562-9c7d-ae6f88e6805d" id="5aa66e96-5fdd-4562-9c7d-ae6f88e6805d"> [...cut..] </vm> <job href="/api/jobs/5c2326c7-f25c-432f-8848-c94e80dd173b" id="5c2326c7-f25c-432f-8848-c94e80dd173b"/> <status> <state>complete</state> </status> </action> --------------------------------------------------------- status of task is always "complete" and the operation take some time, few seconds to some tens of seconds (sync behaviour). Any ideas? Thanks in advance, Giulio

2 3

Enabling GlusterFS in production
by Eduardo Ramos 15 Oct '14

15 Oct '14

Hi friends! I have just upgraded from 3.2 to 3.4 and it was very nice. Everything ok! Now I want to enable Gluster Service to my cluster in production. I think follow this steps: * Install manually vdsm-gluster * Restart supervdsm and vdsmd * Enable Gluster Service in my Cluster. Could it cause any problem or downtime? Thanks.

2 1

Vlans and subnets
by Demeter Tibor 15 Oct '14

15 Oct '14

------=_Part_650053_579042154.1413309486419 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Hello all, I've created a self-hosted ovirt engine with two node, but on the last steps the engine-setup could not attach the node to the "default" cluster. I think it was because I have different services on different vlans. I have vlan100 for glusterfs and vlan101 for vdsm hosts with different subnets. Also, I have an another connection (eth1) with different subnet for internet acces and this is the default route. I would like to separate ovirt services, networks and glusterfs to different vlans/subnets by security reasons. But I don't know what services need to be on same vlan/subnets. My questions/goals - which services need to connect to another ? - Is hosted engine's vm need to be on same network with gluster ? - Is hosted engine's vm need to be on same network with vdsm hosts? - I want to separate ovirt/gluster networks from ovirt VM-s/public internet access - I don't want to extra traffic on router Does anyone experience with this way? Thanks in advance Tibor ------=_Part_650053_579042154.1413309486419 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html><body><div style=3D"font-family: times new roman, new york, times, se= rif; font-size: 12pt; color: #000000"><div><span style=3D"font-size: 12pt;"= >Hello all,</span></div><div><br></div><div>I've created a self-hosted ovir= t engine with two node, but on the last steps the engine-setup could not at= tach the node to the "default" cluster.</div><div>I think it was because I = have different services on different vlans.</div><div>I  have vlan100 = for glusterfs and vlan101 for vdsm hosts with different subnets. Also, I ha= ve an another connection (eth1) with different subnet for internet acces an= d this is the default route.</div><div><br></div><div><span style=3D"font-s= ize: 12pt;"></span><span style=3D"font-size: 12pt;">I would like to separat= e ovirt services, networks and glusterfs to different vlans/subnets by secu= rity reasons.</span></div><div>But I don't know what services need to be on= same vlan/subnets.</div><div><br></div><div>My questions/goals</div><div><= br></div><div>- which services need to connect to another ?</div><div> </di= v><div>- Is hosted engine's vm need to be on same network with gluster ?&nb= sp;</div><div>- Is hosted engine's vm need to be on same network with vdsm = hosts?</div><div>- I want to separate ovirt/gluster networks from ovirt VM-= s/public internet access</div><div>- <span style=3D"font-size: 12pt;">= I don't want to extra traffic on router </span></div><div><br></div><d= iv>Does anyone experience with this way?</div><div><br></div><div>Thanks in= advance</div><div><br></div><div>Tibor </div><div><br></div><div><br>= </div></div></body></html> ------=_Part_650053_579042154.1413309486419--

2 1

Packet loss
by Kyle Gordon 14 Oct '14

14 Oct '14

Evening all, I'm having major issues with a self hosted setup (oVirt 3.4.4 & CentOS 6.5) on Hetzner, with 2 subnets allocated to the server. Routing to and from the additional IP ranges is working as expected, and I can allocate any of the IPs to a virtual interface on the host and get connectivity, so from what I can tell the network routing side of things is correct. However, when I move any of the allocated IPs to any of the guests, I get 1 ICMP response before complete packet loss. This happens at all times, even with iptables completely unloaded and an Accept policy thoughout. The structure is just a basic bridge called ovirtmgmt that all interfaces are bound to. Guests that are affected can communicate with other guests on either of the two subnets, and also with the host, so the bridge is up and running, but when it comes to any traffic trying to egress by the default route, it comes to a halt. In addition, all the hosts have unique MACs, dmesg shows that the appropriate vnets are entering forwarding states and STP is disabled on the bridge. And just to confuse matters, one guest is working just fine - yet when the settings are replicated to any other host the new host still fails to work. Even if I put the working IP onto another guest, it fails to transfer any data. All I get is constant who-has requests on the host bridge whilst the guest looks for the next-hop. Has anyone experienced this before? Where's next on the list of places to look?! Regards Kyle -- Kyle Gordon - 2M1DIQ Web: http://lodge.glasgownet.com Jabber/Email/SIP: kyle(a)lodge.glasgownet.com

1 0

ISO_DOMAIN as iscsi storage type
by ml ml 14 Oct '14

14 Oct '14

Hello List, i am very unhappy with NFS for my ISO_DOMAIN domain. Can i simply remove my ISO_DOMAIN via NFS and add a ISO_DOMAIN via isci? My Setup: http://oi60.tinypic.com/2cmu7nm.jpg I once ran into a pitfall where i was not able to add a ISO DOMAIN anymore, so i would rather ask in advance if i can remove the ISO_DOMAIN and simply add it again via iscsi. Do i need to take care of anything here? Thanks, Mario

2 1