self hosted engine mac range and possible conflict
Hello, suppose I have an environment with self hosted engine environment in 4.1 How is it generated the hw address of the virtual nic of the engine vm when I deploy using cockpit? Suppose on a lan I already have a self hosted engine environment and I'm going to create another one on the same lan, is there any risck of hw address collision for the engine? I know that from inside web admin console I can then set the hw address ranges for the VMs, but what about the mac of the engine vm itself? At the moment I have only lan access and so crosschecked an RHV 4.1 environment and I see that the engine has hw addr 00:16:3e:7c:65:34 (a lookup seems to associate to "Xensource, Inc.") while the VMs have their hw addr in the range 00:1a:4a:16:01:51 - 00:1a:4a:16:01:e6 (eg the 00:1a:4a:16:01:56) (a lookup seems to associate to "Qumranet Inc.". Does this mean that the engine has a sort of dedicated range? Any risk of collision or is a query on the lan done before assigning it? BTW: probably the range for engine VM hw addr has to be put in Qumranet too...? Thanks in advance, Gianluca
On Wed, May 9, 2018 at 4:21 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
Hello, suppose I have an environment with self hosted engine environment in 4.1
How is it generated the hw address of the virtual nic of the engine vm when I deploy using cockpit?
Suppose on a lan I already have a self hosted engine environment and I'm going to create another one on the same lan, is there any risck of hw address collision for the engine?
I know that from inside web admin console I can then set the hw address ranges for the VMs, but what about the mac of the engine vm itself?
At the moment I have only lan access and so crosschecked an RHV 4.1 environment and I see that the engine has hw addr 00:16:3e:7c:65:34 (a lookup seems to associate to "Xensource, Inc.") while the VMs have their hw addr in the range 00:1a:4a:16:01:51 - 00:1a:4a:16:01:e6 (eg the 00:1a:4a:16:01:56) (a lookup seems to associate to "Qumranet Inc.".
Does this mean that the engine has a sort of dedicated range? Any risk of collision or is a query on the lan done before assigning it?
It has been discussed here: https://bugzilla.redhat.com/show_bug.cgi?id=1244502 In a few worlds: the proposed HE MAC address are explicitly on a different range just to minimize collision risks on complex scenarios.
BTW: probably the range for engine VM hw addr has to be put in Qumranet too...?
Thanks in advance,
Gianluca
_______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org
On Wed, May 9, 2018 at 4:32 PM, Simone Tiraboschi <stirabos@redhat.com> wrote:
On Wed, May 9, 2018 at 4:21 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com
wrote:
Hello, suppose I have an environment with self hosted engine environment in 4.1
How is it generated the hw address of the virtual nic of the engine vm when I deploy using cockpit?
Suppose on a lan I already have a self hosted engine environment and I'm going to create another one on the same lan, is there any risck of hw address collision for the engine?
I know that from inside web admin console I can then set the hw address ranges for the VMs, but what about the mac of the engine vm itself?
At the moment I have only lan access and so crosschecked an RHV 4.1 environment and I see that the engine has hw addr 00:16:3e:7c:65:34 (a lookup seems to associate to "Xensource, Inc.") while the VMs have their hw addr in the range 00:1a:4a:16:01:51 - 00:1a:4a:16:01:e6 (eg the 00:1a:4a:16:01:56) (a lookup seems to associate to "Qumranet Inc.".
Does this mean that the engine has a sort of dedicated range? Any risk of collision or is a query on the lan done before assigning it?
It has been discussed here: https://bugzilla.redhat.com/show_bug.cgi?id=1244502
In a few worlds: the proposed HE MAC address are explicitly on a different range just to minimize collision risks on complex scenarios.
Thanks Sandro, Indeed I forgot that as part of the engine deploy (from cockpit but it should be the same from the deprecated command line setup too) there is the explicit option to specify hw address of its vnic. In my case I'm proposed this one: " You may specify a unicast MAC address for the VM or accept a randomly generated default [00:16:3e:15:6a:e5]: " So I should be in the safe place, as I have only one other engine vm and I have no Xen based environments... Gianluca
On Wed, May 9, 2018 at 5:36 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
On Wed, May 9, 2018 at 4:32 PM, Simone Tiraboschi <stirabos@redhat.com> wrote:
On Wed, May 9, 2018 at 4:21 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
Hello, suppose I have an environment with self hosted engine environment in 4.1
How is it generated the hw address of the virtual nic of the engine vm when I deploy using cockpit?
Suppose on a lan I already have a self hosted engine environment and I'm going to create another one on the same lan, is there any risck of hw address collision for the engine?
I know that from inside web admin console I can then set the hw address ranges for the VMs, but what about the mac of the engine vm itself?
At the moment I have only lan access and so crosschecked an RHV 4.1 environment and I see that the engine has hw addr 00:16:3e:7c:65:34 (a lookup seems to associate to "Xensource, Inc.") while the VMs have their hw addr in the range 00:1a:4a:16:01:51 - 00:1a:4a:16:01:e6 (eg the 00:1a:4a:16:01:56) (a lookup seems to associate to "Qumranet Inc.".
Does this mean that the engine has a sort of dedicated range? Any risk of collision or is a query on the lan done before assigning it?
It has been discussed here: https://bugzilla.redhat.com/show_bug.cgi?id=1244502
In a few worlds: the proposed HE MAC address are explicitly on a different range just to minimize collision risks on complex scenarios.
Thanks Sandro, Indeed I forgot that as part of the engine deploy (from cockpit but it should be the same from the deprecated command line setup too) there is the explicit option to specify hw address of its vnic. In my case I'm proposed this one: " You may specify a unicast MAC address for the VM or accept a randomly generated default [00:16:3e:15:6a:e5]: " So I should be in the safe place, as I have only one other engine vm and I have no Xen based environments...
I also filed around the same time: https://bugzilla.redhat.com/show_bug.cgi?id=1516445 Feel free to comment/vote/etc... Best regards, -- Didi
On Thu, May 10, 2018 at 1:59 AM, Yedidyah Bar David <didi@redhat.com> wrote:
On Wed, May 9, 2018 at 5:36 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
On Wed, May 9, 2018 at 4:32 PM, Simone Tiraboschi <stirabos@redhat.com> wrote:
On Wed, May 9, 2018 at 4:21 PM, Gianluca Cecchi <gianluca.cecchi@gmail.com> wrote:
Hello, suppose I have an environment with self hosted engine environment in
4.1
How is it generated the hw address of the virtual nic of the engine vm when I deploy using cockpit?
Suppose on a lan I already have a self hosted engine environment and I'm
going to create another one on the same lan, is there any risck of hw address collision for the engine?
Unfortunately, the MAC address is hard-coded in the most recent release, which will cause collisions if you're deploying a second hosted engine environment on the same LAN. As a workaround, you can edit the hard-coded address in the cockpit UI, or as you're already aware, the CLI version will generate one for you if you use it. I pushed a patch [1] yesterday to correct the issue in cockpit, so the issue should be resolved as of the next release. I also opened a bug [2] to track it, as well. I apologize for the inconvenience. [1] https://gerrit.ovirt.org/#/c/91111/ [2] https://bugzilla.redhat.com/show_bug.cgi?id=1576783
I know that from inside web admin console I can then set the hw address ranges for the VMs, but what about the mac of the engine vm itself?
At the moment I have only lan access and so crosschecked an RHV 4.1 environment and I see that the engine has hw addr 00:16:3e:7c:65:34 (a lookup seems to associate to "Xensource, Inc.") while the VMs have
their hw
addr in the range 00:1a:4a:16:01:51 - 00:1a:4a:16:01:e6 (eg the 00:1a:4a:16:01:56) (a lookup seems to associate to "Qumranet Inc.".
Does this mean that the engine has a sort of dedicated range? Any risk of collision or is a query on the lan done before assigning it?
It has been discussed here: https://bugzilla.redhat.com/show_bug.cgi?id=1244502
In a few worlds: the proposed HE MAC address are explicitly on a different range just to minimize collision risks on complex scenarios.
Thanks Sandro, Indeed I forgot that as part of the engine deploy (from cockpit but it should be the same from the deprecated command line setup too) there is the explicit option to specify hw address of its vnic. In my case I'm proposed this one: " You may specify a unicast MAC address for the VM or accept a randomly generated default [00:16:3e:15:6a:e5]: " So I should be in the safe place, as I have only one other engine vm and I have no Xen based environments...
I also filed around the same time:
https://bugzilla.redhat.com/show_bug.cgi?id=1516445
Feel free to comment/vote/etc...
Best regards, -- Didi _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org
participants (4)
-
Gianluca Cecchi -
Phillip Bailey -
Simone Tiraboschi -
Yedidyah Bar David