[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Ovirt-devel] Managed node NIC management



mark wagner wrote:
Darryl

Not really involved in this discussion previously so may be a naive question(s) but can you key off of the MAC ID of each of the interfaces ?

I realize that this may change if its not integrated on the mobo (and yes even the mobo macid can be changed), but it would seem to handle the overwhelming majority of the cases. If the mac needs to change, treat it as a new card. btw - is there an existing method for dealing with new interfaces ?

I agree. New MAC for a card == totally new interface. New interfaces are handled on each boot, as the node always sends hw information to the server on each boot.

Also (potentially showing more ignorance) I will assume that there may be multiple paths between the managed node and the ovirt server. Are we setting arp_filter=1 in /etc/sysctl.conf ? (we should)

There could possibly be multiple paths between the Node and the Server (we aren't going to prohibit this explicitly). But only a single interface is designated as the iface for the Management LAN and that iface will be the one configured to PXE boot. The additional interfaces brought up for use by VMs for data networks might not even have IP addresses. (In fact I don't think there is a need for them to have addresses)

i.e. Node has 2 NICs. NIC 1 is configured to PXE boot and is on the Management LAN. NIC 2 is configured as a bridge for VMs to use on a VM LAN.

When the node boots, NIC 1 is used to PXE the Node and gets an IP address to talk to the Server. NIC 2 is brought up as a bridge device but doesn't need an IP address. VMs are brought up and use NIC 2 as the bridge and get IPs on the LAN that NIC 2 is connected to.

Now, if for some reason you -do- want to assign an IP to NIC 2, you can (is there a compelling reason to allow this?). And if there is a route to the Server Suite via NIC 2 we don't want the Node using NIC 2 for SS comms, only NIC 1.

Are you saying setting arp_filter=1 would take care of this, if so we should set it as long as it doesn't have any other negative implications.

Also, is the PXE service tied to a particular interface or just the server machine, For instance if the ovirt server and the manage node each have two nics, one on each subnet, do we need to make sure that the PXE request only goes on a certain subnet or is that handled for us?

Each physical node is assumed to have only a single Ethernet interface configured to PXE boot and this interface is the one that is on the Management LAN. This is something that needs to be set up when racking the new box in the datacenter. (i.e. admin boots the box enters the BIOS and picks which cards to PXE boot or not)

Perry


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]