[one-users] Problem with ebtables-kvm script.
Jaime Melis
j.melis at fdi.ucm.es
Mon Dec 14 10:37:30 PST 2009
Hi Gianni,
Try making the following change to the ebtables-kvm script:
58c58
< nets=`virsh dumpxml #{VM_NAME}`
---
> nets=`virsh -c qemu:///system dumpxml #{VM_NAME}`
It will connect directly to the system socket.
Let us know if it works so we can include it in the distribution.
Regards,
Jaime
On Mon, Dec 14, 2009 at 7:20 PM, Jaime Melis <j.melis at fdi.ucm.es> wrote:
> Hi Gianni,
> apparently there's something wrong with your libvirt configuration. Try
> googling the "No vport operation path" error to see if something comes up.
>
> Regards,
> Jaime
>
>
>
> On Mon, Dec 14, 2009 at 3:09 PM, Gianni Dalla Torre <
> gianni.dallatorre at cnaf.infn.it> wrote:
>
>> Hi Jamie,
>>
>> I have no problem connecting to qemu:///system using virsh as oneadmin:
>> oneadmin at omii001:virsh -c qemu:///system
>> Welcome to virsh, the virtualization interactive terminal.
>>
>> Type: 'help' for help with commands
>> 'quit' to quit
>>
>> but I think it is strange that, if I run:
>> oneadmin at omii001:virsh
>> 14:56:19.573: error : No vport operation path found for host1
>> 14:56:19.588: error : No vport operation path found for host0
>> 14:56:19.609: error : No vport operation path found for host2
>> 14:56:19.666: warning : Cannot set group when not running as root
>> Welcome to virsh, the virtualization interactive terminal.
>>
>> Type: 'help' for help with commands
>> 'quit' to quit
>>
>> I don't see this line
>> Connecting to uri: qemu:///system
>> which, instead, I can see if I run the command on other node ivanov:
>>
>> oneadmin at ivanov:virsh
>> Connecting to uri: qemu:///system
>> Welcome to virsh, the virtualization interactive terminal.
>>
>> Type: 'help' for help with commands
>> 'quit' to quit
>>
>> this is the other output, that you request me:
>>
>> oneadmin at omii001$ ls -l /var/run/libvirt/
>> total 8
>> srwxrwx--- 1 root libvirtd 0 Dec 14 10:05 libvirt-sock
>> srwxrwxrwx 1 root libvirtd 0 Dec 14 10:05 libvirt-sock-ro
>> drwxr-xr-x 2 root root 4096 Dec 14 10:05 network
>> drwxr-xr-x 2 root root 4096 Dec 14 09:56 qemu
>>
>> Thank you very much,
>> Gianni.
>>
>>
>>
>> On 14/12/2009 11:49, Jaime Melis wrote:
>>
>> HI Gianni,
>> quick question, can you do
>> virsh -c qemu:///system
>> as oneadmin in omnii001? (explicitly without sudo).
>>
>> If you can't you should add oneadmin to the group libvirtd in omni001. If
>> after that, it still doesn't work, send me the output of:
>> ls -l /var/run/libvirt/
>>
>> regards,
>> Jaime
>>
>> On Fri, Dec 11, 2009 at 4:41 PM, Gianni Dalla Torre <
>> gianni.dallatorre at cnaf.infn.it> wrote:
>>
>>> Hi Jamie,
>>>
>>> I'm still stuck..
>>> The virtual machines deployed by kvm-opennebula driver have this rights:
>>> -rw------- 1 root root 11 Dec 2 15:38 /var/run/libvirt/qemu/one-50.pid
>>> I suppose there is something wrong because with this settings I can't see
>>> the VM as oneadmin, in fact:
>>>
>>> oneamdin at omii001$ virsh list
>>> 16:24:54.622: error : No vport operation path found for host1
>>> 16:24:54.637: error : No vport operation path found for host0
>>> 16:24:54.658: error : No vport operation path found for host2
>>> 16:24:54.715: warning : Cannot set group when not running as root
>>> Id Name State
>>> ----------------------------------
>>>
>>> oneamdin at omii001$ sudo virsh list
>>> Id Name State
>>> ----------------------------------
>>> 4 one-50 running
>>>
>>> Where and by who the VM are deployed?
>>>
>>> This is my report:
>>>
>>> Fri Dec 11 15:38:34 2009 [VMM][D]: Message received: DEPLOY SUCCESS 50
>>> one-50
>>> Fri Dec 11 15:39:01 2009 [VMM][I]: Monitoring VM 50.
>>> Fri Dec 11 15:39:02 2009 [VMM][D]: Message received: POLL SUCCESS 50
>>> STATE=a USEDMEMORY=524288
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 Command
>>> execution fail: '/srv/cloud/one/share/hooks/ebtables-kvm one-50'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 STDERR
>>> follows.
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> 15:37:27.779: error : No vport operation path found for host1
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> 15:37:27.795: error : No vport operation path found for host0
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> 15:37:27.817: error : No vport operation path found for host2
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> 15:37:27.877: warning : Cannot set group when not running as root
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> 15:37:28.064: error : Domain not found: no domain with matching name
>>> 'one-50'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 libvir:
>>> QEMU error : Domain not found: no domain with matching name 'one-50'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 error:
>>> failed to get domain 'one-50'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 error:
>>> Domain not found: no domain with matching name 'one-50'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> /usr/lib/ruby/1.8/rexml/encoding.rb:47: command not found: brctl show
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50
>>> /srv/cloud/one/share/hooks/ebtables-kvm:36:in `get_bridges': undefined
>>> method `each' for nil:Ni
>>> lClass (NoMethodError)
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 from
>>> /srv/cloud/one/share/hooks/ebtables-kvm:50:in `get_interfaces'
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 from
>>> /srv/cloud/one/share/hooks/ebtables-kvm:62
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: LOG - 50 ExitCode: 1
>>> Fri Dec 11 15:39:05 2009 [HKM][D]: Message received: EXECUTE FAILURE 50
>>> ebtables-running
>>>
>>> Thanks in advance for the help.
>>>
>>> Regards,
>>> Gianni
>>>
>>>
>>>
>>> On 30/11/2009 12:04, Jaime Melis wrote:
>>>
>>> Hi Gianni,
>>> I have updated the documentation for the new set of ebtables scripts:
>>>
>>> http://opennebula.org/doku.php?id=documentation:rel1.4:nm
>>>
>>> Regarding your problems, please verify the following things:
>>>
>>> * you have properly set the VM_HOOK configuration in oned.conf,
>>> especially the remote = "yes" option, and restarted one.
>>> * brctl is installed in your worker nodes
>>> * oneadmin user has permissions to execute "sudo ebtables" and "sudo
>>> virsh" in the worker nodes.
>>>
>>> regards,
>>> Jaime
>>>
>>> On Thu, Nov 26, 2009 at 4:58 PM, Gianni Dalla Torre <
>>> gianni.dallatorre at cnaf.infn.it> wrote:
>>>
>>>> Hi Jaime,
>>>>
>>>> thank you for the support, I changed the oned.conf, but I still have
>>>> problems (using argv[1] in the script, was the same to omit start indeed).
>>>> Checking the vm.log seems monitor information has some problems:
>>>>
>>>> Thu Nov 26 14:29:01 2009 [LCM][I]: New VM state is BOOT
>>>> Thu Nov 26 14:29:01 2009 [VMM][I]: Generating deployment file:
>>>> /srv/cloud/one/var/38/deployment.0
>>>> Thu Nov 26 14:29:03 2009 [LCM][I]: New VM state is RUNNING
>>>> Thu Nov 26 14:29:20 2009 [VMM][D]: Monitor Information:
>>>> CPU : -1
>>>> Memory: 524288
>>>> Net_TX: -1
>>>> Net_RX: -1
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: Command execution fail:
>>>> '/srv/cloud/one/share/hooks/ebtables-kvm one-38'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: STDERR follows.
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: 14:29:03.836: error : No vport
>>>> operation path found for host1
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: 14:29:03.851: error : No vport
>>>> operation path found for host0
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: 14:29:03.872: error : No vport
>>>> operation path found for host2
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: 14:29:03.929: warning : Cannot set
>>>> group when not running as root
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: 14:29:04.119: error : Domain not
>>>> found: no domain with matching name 'one-38'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: libvir: QEMU error : Domain not
>>>> found: no domain with matching name 'one-38'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: error: failed to get domain 'one-38'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: error: Domain not found: no domain
>>>> with matching name 'one-38'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]:
>>>> /usr/lib/ruby/1.8/rexml/encoding.rb:47: command not found: brctl show
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]:
>>>> /srv/cloud/one/share/hooks/ebtables-kvm:36:in `get_bridges': undefined
>>>> method `each' for nil:NilClass (NoMethodError)
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: from
>>>> /srv/cloud/one/share/hooks/ebtables-kvm:50:in `get_interfaces'
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: from
>>>> /srv/cloud/one/share/hooks/ebtables-kvm:62
>>>> Thu Nov 26 14:29:33 2009 [HKM][I]: ExitCode: 1
>>>> Thu Nov 26 14:29:33 2009 [HKM][E]: Error executing Hook:
>>>> ebtables-running.
>>>>
>>>> Any other suggestions?
>>>>
>>>> regards,
>>>> Gianni
>>>>
>>>>
>>>> Jaime Melis wrote:
>>>>
>>>> Hello Gianni,
>>>> I'm afraid the documentation section related to the isolation of
>>>> networks has not yet been properly updated (
>>>> http://opennebula.org/doku.php?id=documentation:rel1.4:nm).
>>>>
>>>> Please allow me one or two days to update the relevant section.
>>>>
>>>> In the meantime try with the following confiuration:
>>>>
>>>> The "oned.conf" should be:
>>>>
>>>> arguments = "one-$VMID",
>>>> instead of
>>>> arguments = "start one-$VMID",
>>>>
>>>> so please go ahead and change that in your oned.conf.
>>>>
>>>> As soon as I finish updating the configuration I will let you know.
>>>>
>>>>
>>>> Thanks for your feedback.
>>>>
>>>> regards,
>>>> Jaime
>>>>
>>>>
>>>>
>>>> On Wed, Nov 25, 2009 at 3:02 PM, Gianni Dalla Torre <
>>>> gianni.dallatorre at cnaf.infn.it> wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> I update my installation to the last 1.3.90 openNenbula version, and
>>>>> I'm trying to give some public Ip to my VM. using Bridged networking and the
>>>>> scripts included in the last release.
>>>>>
>>>>> But I have a problem with this script:
>>>>> /srv/cloud/one/share/hooks/ebtables-kvm
>>>>>
>>>>> This is the log I have:
>>>>> Wed Nov 25 12:41:45 2009 [VMM][D]: Message received: DEPLOY SUCCESS 36
>>>>> one-36
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 Command
>>>>> execution fail: '/srv/cloud/one/share/hooks/ebtables-kvm start one-36'
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 STDERR
>>>>> follows.
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 error:
>>>>> failed to get domain 'start'
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 error:
>>>>> Domain not found: no domain with matching name 'start'
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36
>>>>> /usr/lib/ruby/1.8/rexml/encoding.rb:47: command not found: brctl show
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36
>>>>> /srv/cloud/one/share/hooks/ebtables-kvm:36:in `get_bridges': undefined
>>>>> method `each' for nil:NilClass (NoMethodError)
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 from
>>>>> /srv/cloud/one/share/hooks/ebtables-kvm:50:in `get_interfaces'
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 from
>>>>> /srv/cloud/one/share/hooks/ebtables-kvm:62
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: LOG - 36 ExitCode:
>>>>> 1
>>>>> Wed Nov 25 12:41:46 2009 [HKM][D]: Message received: EXECUTE FAILURE 36
>>>>> ebtables-running
>>>>> Wed Nov 25 12:42:10 2009 [VMM][I]: Monitoring VM 36.
>>>>> Wed Nov 25 12:42:11 2009 [VMM][D]: Message received: POLL SUCCESS 36
>>>>> STATE=a USEDMEMORY=524288
>>>>> Wed Nov 25 12:42:19 2009 [InM][D]: Host 4 successfully monitored.
>>>>>
>>>>> I also had a go changing the assignation of argv[0] to argv [1] inside
>>>>> the script, but it didn't work.
>>>>>
>>>>> This are some information about my host configuration:
>>>>>
>>>>> I configured an OpenNebula (v 1.3.90) front-end in my pc "solaris"
>>>>> (Ubuntu 9.04) and installed a cluster node "omii001" (Scientific Linux CERN
>>>>> SLC
>>>>> release 2.6.18-164.6.1.el5 x86_64).
>>>>> I need to access my VM through ssh, so I used the Bridged
>>>>> networking("shared physical device") following this instructions:
>>>>> http://www.opennebula.org/doku.php?id=documentation:rel1.4:nm
>>>>>
>>>>> oneadmin at omii001$ brctl show
>>>>> bridge name bridge id STP enabled interfaces
>>>>> br0 8000.0015c5edbf17 no vnet0
>>>>> eth0
>>>>> virbr0 8000.000000000000 yes
>>>>>
>>>>> Setting ifcfg-eth0 and ifcfg-br0 in this way:
>>>>>
>>>>> oneadmin at omii001$ cat /etc/sysconfig/network-scripts/ifcfg-eth0
>>>>> # Broadcom Corporation NetXtreme II BCM5708 Gigabit Ethernet
>>>>> DEVICE=eth0
>>>>> BOOTPROTO=static
>>>>> DHCPCLASS=
>>>>> HWADDR=00:15:C5:ED:BF:17
>>>>> IPADDR=131.154.100.175
>>>>> NETMASK=255.255.255.0
>>>>> ONBOOT=yes
>>>>> BRIDGE=br0
>>>>>
>>>>> oneadmin at omii001$ cat /etc/sysconfig/network-scripts/ifcfg-br0
>>>>> DEVICE=br0
>>>>> TYPE=Bridge
>>>>> BOOTPROTO=dhcp
>>>>> ONBOOT=yes
>>>>> DELAY=0
>>>>>
>>>>> And this is the template of the VM:
>>>>>
>>>>> #---------------------------------------
>>>>> # VM definition example
>>>>> #---------------------------------------
>>>>> NAME = two
>>>>> CPU = 1 MEMORY = 512
>>>>> # --- kernel & boot device ---
>>>>> OS = [ boot = hd]
>>>>> # --- 1 disks ---
>>>>> DISK = [
>>>>> source = "/srv/cloud/one/images/debian5.img",
>>>>> target = "vda",
>>>>> readonly = "no" ]
>>>>> # --- I/O ---
>>>>> GRAPHICS= [
>>>>> type = "vnc",
>>>>> listen = "127.0.0.1",
>>>>> port = "1"]
>>>>> # --- 2 NIC ---
>>>>> #NIC = [ NETWORK = "OpenNebulaLAN",
>>>>> # BRIDGE = "wlan0" ]
>>>>> NIC = [ NETWORK = "PublicVLAN",
>>>>> BRIDGE = "br0" ]
>>>>> SAVE = yes
>>>>>
>>>>> Some other information about my configuration are in the previous mail:
>>>>>
>>>>> http://lists.opennebula.org/pipermail/users-opennebula.org/2009-November/001202.html
>>>>>
>>>>> If someone have some advice, I would really appreciate it.
>>>>>
>>>>> Best regards,
>>>>> Gianni.
>>>>>
>>>>>
>>>>> |-----|
>>>>> |NOTE:|
>>>>> |-----|
>>>>> I'm not sure if it could be related, but when I compiled opennebula I
>>>>> had problems with some dependencies, but anyway the installation seems to
>>>>> work.
>>>>> ( I'm sure I installed all the the software dependencies listed in the
>>>>> "Platform Notes 1.4" page). I also read some other people had the same issue
>>>>> in the past, but I
>>>>> cannot find the solution and neither any package for Ubuntu that
>>>>> satisfies the requirements.
>>>>>
>>>>> oneadmin at solaris:~/one-1.3.90$ <oneadmin at solaris:%7E/one-1.3.90$> sudo
>>>>> scons
>>>>> scons: Reading SConscript files ...
>>>>> Testing recipe: pkg-config
>>>>> Error calling pkg-config xmlrpc_server_abyss++ --static --libs
>>>>> Testing recipe: xmlrpc-c-config
>>>>> g++ .xmlrpc_test/xmlrpc_test.a7469a.cc -o
>>>>> .xmlrpc_test/xmlrpc_test.a7469a.cc.out -L/usr/lib -lxmlrpc_server_abyss++
>>>>> -lxmlrpc_server++ -lxmlrpc_server_abyss -lxmlrpc_server -lxmlrpc_abyss
>>>>> -lpthread -lxmlrpc++ -lxmlrpc -lxmlrpc_util -lxmlrpc_xmlparse
>>>>> -lxmlrpc_xmltok -I/usr/include 1>>.xmlrpc_test/xmlrpc_test.a7469a.cc.log
>>>>> 2>&1
>>>>>
>>>>> Testing recipe: pkg-config
>>>>> Error calling pkg-config xmlrpc_client++ xmlrpc++ --libs
>>>>> Testing recipe: xmlrpc-c-config
>>>>> g++ .xmlrpc_test/xmlrpc_test.2781fc.cc -o
>>>>> .xmlrpc_test/xmlrpc_test.2781fc.cc.out -L/usr/lib -lxmlrpc_client++ -lcurl
>>>>> -lgssapi_krb5 -lxmlrpc_client -lxmlrpc++ -lxmlrpc -lxmlrpc_util
>>>>> -lxmlrpc_xmlparse -lxmlrpc_xmltok -Wl,--rpath -Wl,/usr/lib -I/usr/include
>>>>> 1>>.xmlrpc_test/xmlrpc_test.2781fc.cc.log 2>&1
>>>>>
>>>>> Testing recipe: pkg-config
>>>>> Error calling pkg-config xmlrpc_server_abyss++ --static --libs
>>>>> Testing recipe: xmlrpc-c-config
>>>>> g++ .xmlrpc_test/xmlrpc_test.6d95be.cc -o
>>>>> .xmlrpc_test/xmlrpc_test.6d95be.cc.out -L/usr/lib -lxmlrpc_server_abyss++
>>>>> -lxmlrpc_server++ -lxmlrpc_server_abyss -lxmlrpc_server -lxmlrpc_abyss
>>>>> -lpthread -lxmlrpc++ -lxmlrpc -lxmlrpc_util -lxmlrpc_xmlparse
>>>>> -lxmlrpc_xmltok -I/usr/include 1>>.xmlrpc_test/xmlrpc_test.6d95be.cc.log
>>>>> 2>&1
>>>>>
>>>>> Testing recipe: pkg-config
>>>>> Error calling pkg-config xmlrpc_client++ xmlrpc++ --libs
>>>>> Testing recipe: xmlrpc-c-config
>>>>> g++ .xmlrpc_test/xmlrpc_test.ce42a1.cc -o
>>>>> .xmlrpc_test/xmlrpc_test.ce42a1.cc.out -L/usr/lib -lxmlrpc_client++ -lcurl
>>>>> -lgssapi_krb5 -lxmlrpc_client -lxmlrpc++ -lxmlrpc -lxmlrpc_util
>>>>> -lxmlrpc_xmlparse -lxmlrpc_xmltok -Wl,--rpath -Wl,/usr/lib -I/usr/include
>>>>> 1>>.xmlrpc_test/xmlrpc_test.ce42a1.cc.log 2>&1
>>>>>
>>>>> scons: done reading SConscript files.
>>>>> scons: Building targets ...
>>>>> scons: `.' is up to date.
>>>>> scons: done building targets.
>>>>>
>>>>> --
>>>>> Gianni Dalla Torre
>>>>>
>>>>
>>>>
>>>> --
>>>> Gianni Dalla Torre
>>>> Computer Science Engineering
>>>>
>>>> INFN -CNAF
>>>> Via Ranzani 13/2
>>>> 40127 Bologna, Italy
>>>> Phone: +39 051 6092913
>>>> Fax: +39 051 6092916
>>>> mail: dallatorre&cnaf.infn.it
>>>>
>>>>
>>>
>>
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opennebula.org/pipermail/users-opennebula.org/attachments/20091214/c9294964/attachment-0001.htm>
More information about the Users
mailing list