[Users] Node register bug work around
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663). However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install? I realise the answer might be to hold on for the fix to be commited but if there is an easy way then that would be good.
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago.
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new issue. The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently.
I realise the answer might be to hold on for the fix to be commited but if there is an easy way then that would be good.
Answer should be to use the latest engine source now that the patches are merged. Mike
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
On 6 February 2012 13:18, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago. Ah ok, will there be a new build of the ISO in the releases folder soon?
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new issue. The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently. I have tried to just register the node from the web interface but I get the following in my log file (/var/log/ovirt-engine) and it is put into non-responsive state:
2012-02-06 14:22:23,562 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (pool-5-thread-48) Running command: ActivateVdsCommand internal: false. Entities affected : ID: b0a98c2e-50c6-11e1-9b27-33a2d18354cc Type: VDS 2012-02-06 14:22:23,564 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (pool-5-thread-48) START, SetVdsStatusVDSCommand(vdsId = b0a98c2e-50c6-11e1-9b27-33a2d18354cc, status=Unassigned, nonOperationalReason=NONE), log id: 2e0506d3 2012-02-06 14:22:23,569 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (pool-5-thread-48) FINISH, SetVdsStatusVDSCommand, log id: 2e0506d3 2012-02-06 14:22:23,585 INFO [org.ovirt.engine.core.vdsbroker.ActivateVdsVDSCommand] (pool-5-thread-48) START, ActivateVdsVDSCommand(vdsId = b0a98c2e-50c6-11e1-9b27-33a2d18354cc), log id: 2033a304 2012-02-06 14:22:23,590 ERROR [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) VDS::handleNetworkException Server failed to respond, vds_id = b0a98c2e-50c6-11e1-9b27-33a2d18354cc, vds_name = jersey, error = VDSNetworkException: 2012-02-06 14:22:23,593 INFO [org.ovirt.engine.core.bll.VdsEventListener] (pool-5-thread-47) ResourceManager::vdsNotResponding entered for Host b0a98c2e-50c6-11e1-9b27-33a2d18354cc, ###.###.###.### 2012-02-06 14:22:23,595 INFO [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) ResourceManager::activateVds - failed to get VDS = b0a98c2e-50c6-11e1-9b27-33a2d18354cc capabilities with error: VDSNetworkException: . 2012-02-06 14:22:23,595 INFO [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) ResourceManager::activateVds - failed to activate VDS = b0a98c2e-50c6-11e1-9b27-33a2d18354cc 2012-02-06 14:22:23,597 ERROR [org.ovirt.engine.core.bll.VdsNotRespondingTreatmentCommand] (pool-5-thread-47) Failed to run Fence script on vds:jersey, VMs moved to UnKnown instead. 2012-02-06 14:22:23,599 INFO [org.ovirt.engine.core.bll.HandleVdsCpuFlagsOrClusterChangedCommand] (pool-5-thread-48) Running command: HandleVdsCpuFlagsOrClusterChangedCommand internal: true. Entities affected : ID: b0a98c2e-50c6-11e1-9b27-33a2d18354cc Type: VDS 2012-02-06 14:22:23,603 WARN [org.ovirt.engine.core.bll.HandleVdsVersionCommand] (pool-5-thread-48) CanDoAction of action HandleVdsVersion failed. Reasons:VDS_CANNOT_CHECK_VERSION_HOST_NON_RESPONSIVE 2012-02-06 14:22:23,603 INFO [org.ovirt.engine.core.vdsbroker.ActivateVdsVDSCommand] (pool-5-thread-48) FINISH, ActivateVdsVDSCommand, log id: 2033a304 I can ssh both directions between the machines.
On Mon, 2012-02-06 at 14:42 +0000, Andrew Dunlop wrote:
On 6 February 2012 13:18, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago. Ah ok, will there be a new build of the ISO in the releases folder soon?
This is an engine fix, not a node fix. You need new engine rpms, not a new iso.
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new issue. The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently. I have tried to just register the node from the web interface but I get the following in my log file (/var/log/ovirt-engine) and it is put into non-responsive state:
This looks like it is a problem with VDSM starting. I *think* it's probably related to a bug in iscsi-initiator-utils [1]. I'm watching this bug and, when fixed, will generate a new iso with the fix. Mike [1]https://bugzilla.redhat.com/show_bug.cgi?id=786174
2012-02-06 14:22:23,562 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (pool-5-thread-48) Running command: ActivateVdsCommand internal: false. Entities affected : ID: b0a98c2e-50c6-11e1-9b27-33a2d18354cc Type: VDS 2012-02-06 14:22:23,564 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (pool-5-thread-48) START, SetVdsStatusVDSCommand(vdsId = b0a98c2e-50c6-11e1-9b27-33a2d18354cc, status=Unassigned, nonOperationalReason=NONE), log id: 2e0506d3 2012-02-06 14:22:23,569 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (pool-5-thread-48) FINISH, SetVdsStatusVDSCommand, log id: 2e0506d3 2012-02-06 14:22:23,585 INFO [org.ovirt.engine.core.vdsbroker.ActivateVdsVDSCommand] (pool-5-thread-48) START, ActivateVdsVDSCommand(vdsId = b0a98c2e-50c6-11e1-9b27-33a2d18354cc), log id: 2033a304 2012-02-06 14:22:23,590 ERROR [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) VDS::handleNetworkException Server failed to respond, vds_id = b0a98c2e-50c6-11e1-9b27-33a2d18354cc, vds_name = jersey, error = VDSNetworkException: 2012-02-06 14:22:23,593 INFO [org.ovirt.engine.core.bll.VdsEventListener] (pool-5-thread-47) ResourceManager::vdsNotResponding entered for Host b0a98c2e-50c6-11e1-9b27-33a2d18354cc, ###.###.###.### 2012-02-06 14:22:23,595 INFO [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) ResourceManager::activateVds - failed to get VDS = b0a98c2e-50c6-11e1-9b27-33a2d18354cc capabilities with error: VDSNetworkException: . 2012-02-06 14:22:23,595 INFO [org.ovirt.engine.core.vdsbroker.VdsManager] (pool-5-thread-48) ResourceManager::activateVds - failed to activate VDS = b0a98c2e-50c6-11e1-9b27-33a2d18354cc 2012-02-06 14:22:23,597 ERROR [org.ovirt.engine.core.bll.VdsNotRespondingTreatmentCommand] (pool-5-thread-47) Failed to run Fence script on vds:jersey, VMs moved to UnKnown instead. 2012-02-06 14:22:23,599 INFO [org.ovirt.engine.core.bll.HandleVdsCpuFlagsOrClusterChangedCommand] (pool-5-thread-48) Running command: HandleVdsCpuFlagsOrClusterChangedCommand internal: true. Entities affected : ID: b0a98c2e-50c6-11e1-9b27-33a2d18354cc Type: VDS 2012-02-06 14:22:23,603 WARN [org.ovirt.engine.core.bll.HandleVdsVersionCommand] (pool-5-thread-48) CanDoAction of action HandleVdsVersion failed. Reasons:VDS_CANNOT_CHECK_VERSION_HOST_NON_RESPONSIVE 2012-02-06 14:22:23,603 INFO [org.ovirt.engine.core.vdsbroker.ActivateVdsVDSCommand] (pool-5-thread-48) FINISH, ActivateVdsVDSCommand, log id: 2033a304
I can ssh both directions between the machines. _______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
On 6 February 2012 15:03, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 14:42 +0000, Andrew Dunlop wrote:
On 6 February 2012 13:18, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago. Ah ok, will there be a new build of the ISO in the releases folder soon?
This is an engine fix, not a node fix. You need new engine rpms, not a new iso. Ah yes sorry.
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new issue. The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently. I have tried to just register the node from the web interface but I get the following in my log file (/var/log/ovirt-engine) and it is put into non-responsive state:
This looks like it is a problem with VDSM starting. I *think* it's probably related to a bug in iscsi-initiator-utils [1]. I'm watching this bug and, when fixed, will generate a new iso with the fix.
Mike
Looks like that's what it is. Thanks
On Monday, 6 February 2012, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 14:42 +0000, Andrew Dunlop wrote:
On 6 February 2012 13:18, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago. Ah ok, will there be a new build of the ISO in the releases folder soon?
This is an engine fix, not a node fix. You need new engine rpms, not a new iso.
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new
issue.
The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently. I have tried to just register the node from the web interface but I get the following in my log file (/var/log/ovirt-engine) and it is put into non-responsive state:
This looks like it is a problem with VDSM starting. I *think* it's probably related to a bug in iscsi-initiator-utils [1]. I'm watching this bug and, when fixed, will generate a new iso with the fix.
Mike
[1]https://bugzilla.redhat.com/show_bug.cgi?id=786174 Looks like that is indeed the problem. Thanks.
--Apple-Mail=_8F87C63F-E46F-4F59-8240-1144DE497B7C Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=iso-8859-1 On 6 Feb 2012, at 18:31, Andrew Dunlop wrote: On Monday, 6 February 2012, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 14:42 +0000, Andrew Dunlop wrote:
On 6 February 2012 13:18, Mike Burns <mburns@redhat.com> wrote:
On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop wrote:
I realise there is a bug open for a problem where nodes can't register with engine (https://bugzilla.redhat.com/show_bug.cgi?id=782663 ).
The fix actually was just pushed a few hours ago. Ah ok, will there be a new build of the ISO in the releases folder soon?
This is an engine fix, not a node fix. You need new engine rpms, not a new iso.
However is there any way to easily work around this problem for now. If I add it from the UI it will try to install the node rather than just registering a pre installed from iso node. Is there anyway of registering it from the management machine without it trying to install?
This actually should work correctly. And if not, then it's a new issue. The engine UI has logic in it to handle when it detects that the host is ovirt-node. This has worked in previous versions of engine/node, but I haven't tested it recently. I have tried to just register the node from the web interface but I get the following in my log file (/var/log/ovirt-engine) and it is put into non-responsive state:
This looks like it is a problem with VDSM starting. I *think* it's probably related to a bug in iscsi-initiator-utils [1]. I'm watching this bug and, when fixed, will generate a new iso with the fix.
Mike
[1]https://bugzilla.redhat.com/show_bug.cgi?id=786174 Looks like that is indeed the problem. Thanks. --Apple-Mail=_8F87C63F-E46F-4F59-8240-1144DE497B7C Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=iso-8859-1
<html><head></head><body style=3D"word-wrap: break-word; = -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; = "><br><div><div>On 6 Feb 2012, at 18:31, Andrew Dunlop = wrote:</div></div>On Monday, 6 February 2012, Mike Burns <<a = href=3D"mailto:mburns@redhat.com">mburns@redhat.com</a>> = wrote:<br>> On Mon, 2012-02-06 at 14:42 +0000, Andrew Dunlop = wrote:<br>>> On 6 February 2012 13:18, Mike Burns <<a = href=3D"mailto:mburns@redhat.com">mburns@redhat.com</a>> = wrote:<br>>> > On Mon, 2012-02-06 at 11:12 +0000, Andrew Dunlop = wrote:<br>>> >> I realise there is a bug open for a problem = where nodes can't register<br>>> >> with engine (<a = href=3D"https://bugzilla.redhat.com/show_bug.cgi?id=3D782663">https://bugz= illa.redhat.com/show_bug.cgi?id=3D782663</a> ).<br>>> = ><br>>> > The fix actually was just pushed a few hours = ago.<br>>> Ah ok, will there be a new build of the ISO in the = releases folder soon?<br>><br>> This is an engine fix, not a node = fix. You need new engine rpms, not a<br>> new = iso.<br>><br>>><br>>> >> However is there any way = to easily work around this problem for now.<br>>> >> If I = add it from the UI it will try to install the node rather = than<br>>> >> just registering a pre installed from iso = node. Is there anyway of<br>>> >> registering it from the = management machine without it trying to<br>>> >> = install?<br>>> ><br>>> > This actually should work = correctly. And if not, then it's a new issue.<br>>> > The = engine UI has logic in it to handle when it detects that the host = is<br>>> > ovirt-node. This has worked in previous = versions of engine/node, but I<br>>> > haven't tested it = recently.<br>>> I have tried to just register the node from the = web interface but I<br>>> get the following in my log file = (/var/log/ovirt-engine) and it is put<br>>> into non-responsive = state:<br>><br>> This looks like it is a problem with VDSM = starting. I *think* it's<br>> probably related to a bug in = iscsi-initiator-utils [1]. I'm watching<br>> this bug and, when = fixed, will generate a new iso with the fix.<br>><br>> = Mike<br>><br>> [1]<a = href=3D"https://bugzilla.redhat.com/show_bug.cgi?id=3D786174">https://bugz= illa.redhat.com/show_bug.cgi?id=3D786174</a><br>Looks like that is = indeed the problem. Thanks.</body></html>= --Apple-Mail=_8F87C63F-E46F-4F59-8240-1144DE497B7C--
participants (4)
-
Andrew Dunlop -
Andrew Dunlop -
Andrew James Dunlop
-
Mike Burns