[Users] Compliance problem with ovirt 3.3 on Fedora 19
I attempted to setup and all-in-one solution on Fedora 19. I ran through the install which made a default data center and cluster. I created a host with 127.0.0.1 and then attempted to add the bridge that the test day documentation states does not get handled properly. After many reboots and disabling NetworkManager.service and enabling network.service as well as manually editing the /etc/sysconfig/network-scripts/ifcfg-* files I had a network connection and reduced the "errors" that seem to be keeping my host in the Nonfunctional state to: Host Localhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} Recommendations of something I may be doing wrong? Has running 3.3 on F19 been a bad choice on my part? Bug?
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Monday, August 26, 2013 11:32:02 PM Subject: [Users] Compliance problem with ovirt 3.3 on Fedora 19
I attempted to setup and all-in-one solution on Fedora 19. I ran through the install which made a default data center and cluster. I created a host with 127.0.0.1 and then attempted to add the bridge that the test day documentation states does not get handled properly. After many reboots and disabling NetworkManager.service and enabling network.service as well as manually editing the /etc/sysconfig/network-scripts/ifcfg-* files I had a network connection and reduced the "errors" that seem to be keeping my host in the Nonfunctional state to:
Host Localhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]}
Recommendations of something I may be doing wrong? Has running 3.3 on F19 been a bad choice on my part? Bug?
The engine does not support localhost (loopback) as an address for network management. You have to use a real interface with real address. You can also use fake interface to do so, such as dummy. modprobe dummy ifconfig dummy0 10.0.0.1 netmask 255.255.255.0 Alon
The engine does not support localhost (loopback) as an address for network management.
You have to use a real interface with real address. I did also try the systems "real" IP and it did not seem to change anything with regards to:
The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]}
You can also use fake interface to do so, such as dummy.
modprobe dummy ifconfig dummy0 10.0.0.1 netmask 255.255.255.0
I'll keep that in mind. May be useful for other things I want to do.
Alon
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Monday, August 26, 2013 11:52:15 PM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
The engine does not support localhost (loopback) as an address for network management.
You have to use a real interface with real address. I did also try the systems "real" IP and it did not seem to change anything with regards to:
The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]}
Logs will be nice.
You can also use fake interface to do so, such as dummy.
modprobe dummy ifconfig dummy0 10.0.0.1 netmask 255.255.255.0
I'll keep that in mind. May be useful for other things I want to do.
Alon
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
This is a multi-part message in MIME format. --------------030603060105010905070902 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Snippets of what seems possibly relevant: https://gist.github.com/thomassuckow/995f422cd5c46c6de98c messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant The engine.log shows the error I mentioned before: ... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} ... ... Message: State was set to NonOperational for host myhost --------------030603060105010905070902 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: 7bit <html> <head> <meta content="text/html; charset=UTF-8" http-equiv="Content-Type"> </head> <body bgcolor="#FFFFFF" text="#000000"> Snippets of what seems possibly relevant: <meta http-equiv="content-type" content="text/html; charset=UTF-8"> <a href="https://gist.github.com/thomassuckow/995f422cd5c46c6de98c">https://gist.github.com/thomassuckow/995f422cd5c46c6de98c</a><br> <br> messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant<br> <br> The engine.log shows the error I mentioned before:<br> <br> ... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]}<br> ...<br> ... Message: State was set to NonOperational for host myhost<br> <br> <br> </body> </html> --------------030603060105010905070902--
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Tuesday, August 27, 2013 1:48:33 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
Snippets of what seems possibly relevant: https://gist.github.com/thomassuckow/995f422cd5c46c6de98c
messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant
The engine.log shows the error I mentioned before:
... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} ... ... Message: State was set to NonOperational for host myhost
For the service issue, I am adding Yaniv. The Hosts emulated machines are <UNKNOWN>, I am adding Yair. As far as I can see the host-deploy is finished OK, just to confirm can you send output of: # ip link # ip addr Thanks! Alon
The Hosts emulated machines are <UNKNOWN>, is probably an error during parsing emulatedMachines that returned from vdsm. doesn't have anything with network IIUC.. so its probably new bug in parsing the list. please run "vdsClient -s 0 getVdsCaps | grep emulatedMachines" on host and send it here. For the second service issue, its quite obvious -> ksmtuned couldn't start with systemd nor initctl, try to start it manually, maybe you just missing 'qemu-kvm' package somehow. Let me know the outputs and I'll try to figure it more deeply. Thanks, Yaniv Bronhaim. ----- Original Message -----
From: "Alon Bar-Lev" <alonbl@redhat.com> To: "Thomas Suckow" <thomas.suckow@pnnl.gov>, "Yaniv Bronheim" <ybronhei@redhat.com>, "Yair Zaslavsky" <yzaslavs@redhat.com> Cc: users@ovirt.org Sent: Tuesday, August 27, 2013 8:55:06 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Tuesday, August 27, 2013 1:48:33 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
Snippets of what seems possibly relevant: https://gist.github.com/thomassuckow/995f422cd5c46c6de98c
messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant
The engine.log shows the error I mentioned before:
... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} ... ... Message: State was set to NonOperational for host myhost
For the service issue, I am adding Yaniv. The Hosts emulated machines are <UNKNOWN>, I am adding Yair.
As far as I can see the host-deploy is finished OK, just to confirm can you send output of: # ip link # ip addr
Thanks! Alon
#vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = [] qemu-kvm is installed, though the ksm package was not. i have now installed it. Interesting lines in messages: vdsm vds WARNING Unable to load the json rpc server module. Please make sure it is installed Aug 27 10:41:06 we13196 kernel: [ 325.223488] device-mapper: table: 253:2: multipath: error getting device Aug 27 10:41:06 we13196 kernel: [ 325.223496] device-mapper: ioctl: error adding target to table Aug 27 10:41:06 we13196 kernel: [ 325.224890] device-mapper: table: 253:2: multipath: error getting device Aug 27 10:41:06 we13196 kernel: [ 325.224898] device-mapper: ioctl: error adding target to table extra getVdsCaps info: operatingSystem = {'name': 'Fedora', 'release': '2', 'version': '19'} packages2 = {'kernel': {'buildtime': 1371004778.0, 'release': '301.fc19.x86_64', 'version': '3.9.5'}, 'libvirt': {'buildtime': 1375400611, 'release': '1.fc19', 'version': '1.0.5.5'}, 'mom': {'buildtime': 1375215820, 'release': '3.fc19', 'version': '0.3.2'}, 'qemu-img': {'buildtime': 1376836471, 'release': '7.fc19', 'version': '1.4.2'}, 'qemu-kvm': {'buildtime': 1376836471, 'release': '7.fc19', 'version': '1.4.2'}, 'spice-server': {'buildtime': 1369154734, 'release': '1.fc19', 'version': '0.12.3'}, 'vdsm': {'buildtime': 1377187393, 'release': '2.fc19', 'version': '4.12.0'}} reservedMem = '321' software_revision = '2' software_version = '4.12' supportedENGINEs = ['3.0', '3.1', '3.2', '3.3'] supportedProtocols = ['2.2', '2.3'] uuid = '...631' version_name = 'Snow Man' vlans = {} vmTypes = ['kvm'] Other things of note: It seems libvirt is starting before the network is ready. /var/log/libvirt/libvirtd.log 2013-08-27 10:36:15.222+0000: 1161: error : virNetSocketNewListenTCP:240 : Unable to resolve address '0.0.0.0' service '16514': Address family for hostname not supported 2013-08-27 10:36:15.222+0000: 1161: info : virNetlinkEventServiceStopAll:422 : stopping all netlink event services Which results in VDSM failing with: Aug 27 03:36:26 we13196 systemd-vdsmd[815]: libvirt.libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory Aug 27 03:36:26 we13196 systemd-vdsmd[815]: vdsm: Failed to define network filters on libvirt[FAILED] Aug 27 03:36:26 we13196 systemd[1]: vdsmd.service: control process exited, code=exited status=1 Aug 27 03:36:26 we13196 systemd[1]: Failed to start Virtual Desktop Server Manager. Aug 27 03:36:26 we13196 systemd[1]: Unit vdsmd.service entered failed state. Clicking activate in the ovirt UI however restarts both services and we end up back at the emulatedMachines = [] issue. Probably not related as I am not seeing other errors in logs, but never know. -- Thomas ________________________________________ From: Yaniv Bronheim [ybronhei@redhat.com] Sent: Tuesday, August 27, 2013 9:31 AM To: Alon Bar-Lev Cc: Suckow, Thomas J; Yair Zaslavsky; users@ovirt.org Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19 The Hosts emulated machines are <UNKNOWN>, is probably an error during parsing emulatedMachines that returned from vdsm. doesn't have anything with network IIUC.. so its probably new bug in parsing the list. please run "vdsClient -s 0 getVdsCaps | grep emulatedMachines" on host and send it here. For the second service issue, its quite obvious -> ksmtuned couldn't start with systemd nor initctl, try to start it manually, maybe you just missing 'qemu-kvm' package somehow. Let me know the outputs and I'll try to figure it more deeply. Thanks, Yaniv Bronhaim. ----- Original Message -----
From: "Alon Bar-Lev" <alonbl@redhat.com> To: "Thomas Suckow" <thomas.suckow@pnnl.gov>, "Yaniv Bronheim" <ybronhei@redhat.com>, "Yair Zaslavsky" <yzaslavs@redhat.com> Cc: users@ovirt.org Sent: Tuesday, August 27, 2013 8:55:06 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Tuesday, August 27, 2013 1:48:33 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
Snippets of what seems possibly relevant: https://gist.github.com/thomassuckow/995f422cd5c46c6de98c
messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant
The engine.log shows the error I mentioned before:
... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} ... ... Message: State was set to NonOperational for host myhost
For the service issue, I am adding Yaniv. The Hosts emulated machines are <UNKNOWN>, I am adding Yair.
As far as I can see the host-deploy is finished OK, just to confirm can you send output of: # ip link # ip addr
Thanks! Alon
Did this get resolved? Is there a BZ? (I can't find one.) I am also seeing "The Hosts emulated machines are <UNKNOWN>" on yesterday's nightly build. The host I'm trying to add is a VM. [host2 $] vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = [] Greg ----- Original Message -----
From: "Thomas J Suckow" <Thomas.Suckow@pnnl.gov> To: "Yaniv Bronheim" <ybronhei@redhat.com>, "Alon Bar-Lev" <alonbl@redhat.com> Cc: users@ovirt.org Sent: Tuesday, August 27, 2013 2:03:26 PM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
#vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = []
qemu-kvm is installed, though the ksm package was not. i have now installed it.
Interesting lines in messages:
vdsm vds WARNING Unable to load the json rpc server module. Please make sure it is installed Aug 27 10:41:06 we13196 kernel: [ 325.223488] device-mapper: table: 253:2: multipath: error getting device Aug 27 10:41:06 we13196 kernel: [ 325.223496] device-mapper: ioctl: error adding target to table Aug 27 10:41:06 we13196 kernel: [ 325.224890] device-mapper: table: 253:2: multipath: error getting device Aug 27 10:41:06 we13196 kernel: [ 325.224898] device-mapper: ioctl: error adding target to table
extra getVdsCaps info:
operatingSystem = {'name': 'Fedora', 'release': '2', 'version': '19'} packages2 = {'kernel': {'buildtime': 1371004778.0, 'release': '301.fc19.x86_64', 'version': '3.9.5'}, 'libvirt': {'buildtime': 1375400611, 'release': '1.fc19', 'version': '1.0.5.5'}, 'mom': {'buildtime': 1375215820, 'release': '3.fc19', 'version': '0.3.2'}, 'qemu-img': {'buildtime': 1376836471, 'release': '7.fc19', 'version': '1.4.2'}, 'qemu-kvm': {'buildtime': 1376836471, 'release': '7.fc19', 'version': '1.4.2'}, 'spice-server': {'buildtime': 1369154734, 'release': '1.fc19', 'version': '0.12.3'}, 'vdsm': {'buildtime': 1377187393, 'release': '2.fc19', 'version': '4.12.0'}} reservedMem = '321' software_revision = '2' software_version = '4.12' supportedENGINEs = ['3.0', '3.1', '3.2', '3.3'] supportedProtocols = ['2.2', '2.3'] uuid = '...631' version_name = 'Snow Man' vlans = {} vmTypes = ['kvm']
Other things of note:
It seems libvirt is starting before the network is ready.
/var/log/libvirt/libvirtd.log 2013-08-27 10:36:15.222+0000: 1161: error : virNetSocketNewListenTCP:240 : Unable to resolve address '0.0.0.0' service '16514': Address family for hostname not supported 2013-08-27 10:36:15.222+0000: 1161: info : virNetlinkEventServiceStopAll:422 : stopping all netlink event services
Which results in VDSM failing with:
Aug 27 03:36:26 we13196 systemd-vdsmd[815]: libvirt.libvirtError: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory Aug 27 03:36:26 we13196 systemd-vdsmd[815]: vdsm: Failed to define network filters on libvirt[FAILED] Aug 27 03:36:26 we13196 systemd[1]: vdsmd.service: control process exited, code=exited status=1 Aug 27 03:36:26 we13196 systemd[1]: Failed to start Virtual Desktop Server Manager. Aug 27 03:36:26 we13196 systemd[1]: Unit vdsmd.service entered failed state.
Clicking activate in the ovirt UI however restarts both services and we end up back at the emulatedMachines = [] issue. Probably not related as I am not seeing other errors in logs, but never know.
-- Thomas ________________________________________ From: Yaniv Bronheim [ybronhei@redhat.com] Sent: Tuesday, August 27, 2013 9:31 AM To: Alon Bar-Lev Cc: Suckow, Thomas J; Yair Zaslavsky; users@ovirt.org Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
The Hosts emulated machines are <UNKNOWN>, is probably an error during parsing emulatedMachines that returned from vdsm. doesn't have anything with network IIUC.. so its probably new bug in parsing the list. please run "vdsClient -s 0 getVdsCaps | grep emulatedMachines" on host and send it here.
For the second service issue, its quite obvious -> ksmtuned couldn't start with systemd nor initctl, try to start it manually, maybe you just missing 'qemu-kvm' package somehow.
Let me know the outputs and I'll try to figure it more deeply.
Thanks,
Yaniv Bronhaim.
----- Original Message -----
From: "Alon Bar-Lev" <alonbl@redhat.com> To: "Thomas Suckow" <thomas.suckow@pnnl.gov>, "Yaniv Bronheim" <ybronhei@redhat.com>, "Yair Zaslavsky" <yzaslavs@redhat.com> Cc: users@ovirt.org Sent: Tuesday, August 27, 2013 8:55:06 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Tuesday, August 27, 2013 1:48:33 AM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
Snippets of what seems possibly relevant: https://gist.github.com/thomassuckow/995f422cd5c46c6de98c
messages seems mostly happy with the exception of a crash in "/usr/bin/vdsm-tool" related to ksmtuned, not sure if relevant
The engine.log shows the error I mentioned before:
... Message: Host myhost does not comply with the cluster Default emulated machines. The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]} ... ... Message: State was set to NonOperational for host myhost
For the service issue, I am adding Yaniv. The Hosts emulated machines are <UNKNOWN>, I am adding Yair.
As far as I can see the host-deploy is finished OK, just to confirm can you send output of: # ip link # ip addr
Thanks! Alon
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
On 09/13/2013 04:29 AM, Greg Sheremeta wrote:
Did this get resolved? Is there a BZ? (I can't find one.) I am also seeing "The Hosts emulated machines are <UNKNOWN>" on yesterday's nightly build. The host I'm trying to add is a VM.
[host2 $] vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = []
Greg
I am uncertain if it did, I took down our production system (Low availability requirements) yesterday so I could fiddle with an all-in-one install trial today on another hard disk. I wonder if it is a certain hardware configuration, ours is an old Dell Precision. My previous attempt was "normal" install and then tried to add localhost. I wonder if the all in one would give different results. - Thomas
On 09/13/2013 09:28 AM, Suckow, Thomas J wrote:
On 09/13/2013 04:29 AM, Greg Sheremeta wrote:
Did this get resolved? Is there a BZ? (I can't find one.) I am also seeing "The Hosts emulated machines are <UNKNOWN>" on yesterday's nightly build. The host I'm trying to add is a VM.
[host2 $] vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = []
Greg
I am uncertain if it did, I took down our production system (Low availability requirements) yesterday so I could fiddle with an all-in-one install trial today on another hard disk. I wonder if it is a certain hardware configuration, ours is an old Dell Precision.
My previous attempt was "normal" install and then tried to add localhost. I wonder if the all in one would give different results.
- Thomas
I just finally got the All-In-One installed. It seems to be working. I noticed that the all-in-one set the CPU type in the cluster settings, you could try setting that...
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Friday, September 13, 2013 8:14:32 PM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
On 09/13/2013 09:28 AM, Suckow, Thomas J wrote:
On 09/13/2013 04:29 AM, Greg Sheremeta wrote:
Did this get resolved? Is there a BZ? (I can't find one.) I am also seeing "The Hosts emulated machines are <UNKNOWN>" on yesterday's nightly build. The host I'm trying to add is a VM.
[host2 $] vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = []
Greg
I am uncertain if it did, I took down our production system (Low availability requirements) yesterday so I could fiddle with an all-in-one install trial today on another hard disk. I wonder if it is a certain hardware configuration, ours is an old Dell Precision.
My previous attempt was "normal" install and then tried to add localhost. I wonder if the all in one would give different results.
- Thomas
I just finally got the All-In-One installed. It seems to be working. I noticed that the all-in-one set the CPU type in the cluster settings, you could try setting that...
Thanks -- it didn't work. I reported this as a bug. I suspect it's a vdsm bug related to the host being virtualized. But just guessing. https://bugzilla.redhat.com/show_bug.cgi?id=1008243 Greg
----- Original Message -----
From: "Greg Sheremeta" <gshereme@redhat.com> To: "Thomas Suckow" <thomas.suckow@pnnl.gov> Cc: users@ovirt.org Sent: Sunday, September 15, 2013 5:43:42 PM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
----- Original Message -----
From: "Thomas Suckow" <thomas.suckow@pnnl.gov> To: users@ovirt.org Sent: Friday, September 13, 2013 8:14:32 PM Subject: Re: [Users] Compliance problem with ovirt 3.3 on Fedora 19
On 09/13/2013 09:28 AM, Suckow, Thomas J wrote:
On 09/13/2013 04:29 AM, Greg Sheremeta wrote:
Did this get resolved? Is there a BZ? (I can't find one.) I am also seeing "The Hosts emulated machines are <UNKNOWN>" on yesterday's nightly build. The host I'm trying to add is a VM.
[host2 $] vdsClient -s 0 getVdsCaps | grep emulatedMachines emulatedMachines = []
Greg
I am uncertain if it did, I took down our production system (Low availability requirements) yesterday so I could fiddle with an all-in-one install trial today on another hard disk. I wonder if it is a certain hardware configuration, ours is an old Dell Precision.
My previous attempt was "normal" install and then tried to add localhost. I wonder if the all in one would give different results.
- Thomas
I just finally got the All-In-One installed. It seems to be working. I noticed that the all-in-one set the CPU type in the cluster settings, you could try setting that...
Thanks -- it didn't work. I reported this as a bug. I suspect it's a vdsm bug related to the host being virtualized. But just guessing. https://bugzilla.redhat.com/show_bug.cgi?id=1008243
Greg
Following up. Thanks to Assaf Muller for the tip: running a yum update on a host works around the "The Hosts emulated machines are <UNKNOWN> and the cluster is [rhel6.4.0, pc-1.0]}". I guess it's just best to do that on all my new hosts from now on. Not sure if it's officially recommended or documented to do that. Follow BZ1008243 if you're interested in the root cause. Greg
participants (5)
-
Alon Bar-Lev -
Greg Sheremeta -
Suckow, Thomas J -
Thomas Suckow
-
Yaniv Bronheim