ovirtmgmt bridge, hosted engine, and running VMs
This is a multi-part message in MIME format. --------------080303010305060104030604 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Hi folks, I have a 3.5 hosted-engine setup, which was recently upgraded from 3.4. It has five nodes, two of them set up for hosted-engine HA. Initial problem: One of the hosted-engine HVs had a score of 1800, owing to the management bridge (ovirtmgmt) being absent. I put ovirtmgmt bridge back in. The score went to 2400. Great! After that, I could use hosted-engine --set-maintenance --mode=local to force it to move to the other blade. Can't migrate hosted-engine using the oVirt GUI though; I'm sure I've done so before, is that still supposed to work in 3.5? Second problem: Having added the ovirtmgmt bridge, that HV, which was running VMs before, now refuses to accept them. An example from hosted-engine's engine.log, it refuses to put a VM on these hosts: 2015-04-29 17:23:07,593 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl09.networkvideo.com.au (4f26611a-9f44-4832-b9e3-1a06b1d513fc) was filtered out by VAR__FILTERTYPE__INTERNAL *filter Network* 2015-04-29 17:23:07,595 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl07.networkvideo.com.au (fc50be91-3e07-4447-a0d8-bffbda8a07c6) was filtered out by VAR__FILTERTYPE__INTERNAL filter Network I think it must relate to the configuration for either the ovirtmgmt bridge, or the physical interface to which the bridge is connected. I have seen, for example, the need for BOOTPROTO=none to be in the ifcfg-file. So, it seems that filter actually reads config files, and I suspect it's looking for a certain directive and/or syntax. My guess, it's being (too) fussy about syntax. Has anyone else encountered this? I'd be glad to learn more about how that filter works, if someone can point me in the right direction please. The interface configs for the two scenarios are shown below. 1. eth0 without a bridge - like this, I can run VMs on the HV, but hosted-engine won't go there (of course). [root@bl09 network-scripts]# cat ifcfg-eth0 DEVICE=eth0 HWADDR=00:21:5a:48:4e:4a ONBOOT=yes IPADDR=10.0.14.9 NETMASK=255.255.255.0 GATEWAY=10.0.14.254 BOOTPROTO=none MTU=1500 DEFROUTE=yes NM_CONTROLLED=no 2. With the bridge in, as it is now, I can put hosted-engine on it, but can't run other VMs on there. [root@bl09 network-scripts]# cat ifcfg-eth0 DEVICE=eth0 HWADDR=00:21:5a:48:4e:4a BOOTPROTO=none ONBOOT=yes BRIDGE=ovirtmgmt MTU=1500 DEFROUTE=no NM_CONTROLLED=no [root@bl09 network-scripts]# cat ifcfg-ovirtmgmt DEVICE=ovirtmgmt ONBOOT=yes TYPE=Bridge DELAY=0 IPADDR=10.0.14.9 NETMASK=255.255.255.0 GATEWAY=10.0.14.254 BOOTPROTO=static DEFROUTE=yes NM_CONTROLLED=no Any guidance appreciated. -- Thanks, Garry --------------080303010305060104030604 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 8bit <html> <head> <meta http-equiv="content-type" content="text/html; charset=utf-8"> </head> <body bgcolor="#FFFFFF" text="#000000"> Hi folks,<br> <br> I have a 3.5 hosted-engine setup, which was recently upgraded from 3.4. It has five nodes, two of them set up for hosted-engine HA.<br> <br> Initial problem:<br> <br> One of the hosted-engine HVs had a score of 1800, owing to the management bridge (ovirtmgmt) being absent. <br> I put ovirtmgmt bridge back in. The score went to 2400. Great!<br> After that, I could use hosted-engine --set-maintenance --mode=local to force it to move to the other blade.<br> <br> Can't migrate hosted-engine using the oVirt GUI though; I'm sure I've done so before, is that still supposed to work in 3.5?<br> <br> Second problem:<br> <br> Having added the ovirtmgmt bridge, that HV, which was running VMs before, now refuses to accept them.<br> <br> An example from hosted-engine's engine.log, it refuses to put a VM on these hosts:<br> <br> <tt>2015-04-29 17:23:07,593 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl09.networkvideo.com.au (4f26611a-9f44-4832-b9e3-1a06b1d513fc) was filtered out by VAR__FILTERTYPE__INTERNAL <b>filter Network</b></tt><tt><br> </tt><tt>2015-04-29 17:23:07,595 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl07.networkvideo.com.au (fc50be91-3e07-4447-a0d8-bffbda8a07c6) was filtered out by VAR__FILTERTYPE__INTERNAL filter Network</tt><br> <br> I think it must relate to the configuration for either the ovirtmgmt bridge, or the physical interface to which the bridge is connected.<br> I have seen, for example, the need for BOOTPROTO=none to be in the ifcfg-file.<br> So, it seems that filter actually reads config files, and I suspect it's looking for a certain directive and/or syntax. My guess, it's being (too) fussy about syntax.<br> <br> Has anyone else encountered this? I'd be glad to learn more about how that filter works, if someone can point me in the right direction please.<br> <br> The interface configs for the two scenarios are shown below.<br> <br> 1. eth0 without a bridge - like this, I can run VMs on the HV, but hosted-engine won't go there (of course).<br> <br> <tt>[root@bl09 network-scripts]# cat ifcfg-eth0</tt><tt><br> </tt><tt>DEVICE=eth0</tt><tt><br> </tt><tt>HWADDR=00:21:5a:48:4e:4a</tt><tt><br> </tt><tt>ONBOOT=yes</tt><tt><br> </tt><tt>IPADDR=10.0.14.9</tt><tt><br> </tt><tt>NETMASK=255.255.255.0</tt><tt><br> </tt><tt>GATEWAY=10.0.14.254</tt><tt><br> </tt><tt>BOOTPROTO=none</tt><tt><br> </tt><tt>MTU=1500</tt><tt><br> </tt><tt>DEFROUTE=yes</tt><tt><br> </tt><tt>NM_CONTROLLED=no<br> <br> </tt>2. With the bridge in, as it is now, I can put hosted-engine on it, but can't run other VMs on there.<br> <br> <tt>[root@bl09 network-scripts]# cat ifcfg-eth0</tt><tt><br> </tt><tt>DEVICE=eth0</tt><tt><br> </tt><tt>HWADDR=00:21:5a:48:4e:4a</tt><tt><br> </tt><tt>BOOTPROTO=none</tt><tt><br> </tt><tt>ONBOOT=yes</tt><tt><br> </tt><tt>BRIDGE=ovirtmgmt</tt><tt><br> </tt><tt>MTU=1500</tt><tt><br> </tt><tt>DEFROUTE=no</tt><tt><br> </tt><tt>NM_CONTROLLED=no</tt><tt><br> </tt><tt>[root@bl09 network-scripts]# cat ifcfg-ovirtmgmt</tt><tt><br> </tt><tt>DEVICE=ovirtmgmt</tt><tt><br> </tt><tt>ONBOOT=yes</tt><tt><br> </tt><tt>TYPE=Bridge</tt><tt><br> </tt><tt>DELAY=0</tt><tt><br> </tt><tt>IPADDR=10.0.14.9</tt><tt><br> </tt><tt>NETMASK=255.255.255.0</tt><tt><br> </tt><tt>GATEWAY=10.0.14.254</tt><tt><br> </tt><tt>BOOTPROTO=static</tt><tt><br> </tt><tt>DEFROUTE=yes</tt><tt><br> </tt><tt>NM_CONTROLLED=no</tt><br> <br> Any guidance appreciated.<br> <br> -- <br> <div class="moz-signature">Thanks,<br> <br> Garry<br> </div> </body> </html> --------------080303010305060104030604--
Il 29/04/2015 09:55, Garry Tiedemann ha scritto:
Hi folks,
I have a 3.5 hosted-engine setup, which was recently upgraded from 3.4. It has five nodes, two of them set up for hosted-engine HA.
Initial problem:
One of the hosted-engine HVs had a score of 1800, owing to the management bridge (ovirtmgmt) being absent. I put ovirtmgmt bridge back in. The score went to 2400. Great! After that, I could use hosted-engine --set-maintenance --mode=local to force it to move to the other blade.
Can't migrate hosted-engine using the oVirt GUI though; I'm sure I've done so before, is that still supposed to work in 3.5?
AFAIK yes, it should still be supported.
Second problem:
Having added the ovirtmgmt bridge, that HV, which was running VMs before, now refuses to accept them.
An example from hosted-engine's engine.log, it refuses to put a VM on these hosts:
2015-04-29 17:23:07,593 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl09.networkvideo.com.au (4f26611a-9f44-4832-b9e3-1a06b1d513fc) was filtered out by VAR__FILTERTYPE__INTERNAL *filter Network* 2015-04-29 17:23:07,595 INFO [org.ovirt.engine.core.bll.scheduling.SchedulingManager] (ajp--127.0.0.1-8702-2) [4ca60585] Candidate host bl07.networkvideo.com.au (fc50be91-3e07-4447-a0d8-bffbda8a07c6) was filtered out by VAR__FILTERTYPE__INTERNAL filter Network
I think it must relate to the configuration for either the ovirtmgmt bridge, or the physical interface to which the bridge is connected. I have seen, for example, the need for BOOTPROTO=none to be in the ifcfg-file. So, it seems that filter actually reads config files, and I suspect it's looking for a certain directive and/or syntax. My guess, it's being (too) fussy about syntax.
Has anyone else encountered this? I'd be glad to learn more about how that filter works, if someone can point me in the right direction please.
The interface configs for the two scenarios are shown below.
1. eth0 without a bridge - like this, I can run VMs on the HV, but hosted-engine won't go there (of course).
[root@bl09 network-scripts]# cat ifcfg-eth0 DEVICE=eth0 HWADDR=00:21:5a:48:4e:4a ONBOOT=yes IPADDR=10.0.14.9 NETMASK=255.255.255.0 GATEWAY=10.0.14.254 BOOTPROTO=none MTU=1500 DEFROUTE=yes NM_CONTROLLED=no
2. With the bridge in, as it is now, I can put hosted-engine on it, but can't run other VMs on there.
[root@bl09 network-scripts]# cat ifcfg-eth0 DEVICE=eth0 HWADDR=00:21:5a:48:4e:4a BOOTPROTO=none ONBOOT=yes BRIDGE=ovirtmgmt MTU=1500 DEFROUTE=no NM_CONTROLLED=no [root@bl09 network-scripts]# cat ifcfg-ovirtmgmt DEVICE=ovirtmgmt ONBOOT=yes TYPE=Bridge DELAY=0 IPADDR=10.0.14.9 NETMASK=255.255.255.0 GATEWAY=10.0.14.254 BOOTPROTO=static DEFROUTE=yes NM_CONTROLLED=no
Any guidance appreciated.
-- Thanks,
Garry
_______________________________________________ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
-- Sandro Bonazzola Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com
participants (2)
-
Garry Tiedemann -
Sandro Bonazzola