[rhelv6-list] Does RHEL6 KVM PXE actually work?

[Kirby Zhou]

Sorry I do not think you get the point.
>From my screen shot, you can find that: the file gpxelinux.0 is already
downloaded, so the bridge should be OK.

Finally I have found that the problem is virtio, if take rtl8139 as the
network model, it works with PXE.

Regards,
   Kirby Zhou    
   from   SOHU-RD   +86-10-6272-8261

-----Original Message-----
From: rhelv6-list-bounces at redhat.com [mailto:rhelv6-list-bounces at redhat.com]
On Behalf Of Phil Meyer
Sent: Wednesday, March 16, 2011 10:36 PM
To: rhelv6-list at redhat.com
Subject: Re: [rhelv6-list] Does RHEL6 KVM PXE actually work?

On 03/16/2011 04:51 AM, Kirby Zhou wrote:
>
> Does RHEL6 KVM PXE actually work?
>
>


Your example shows a bridge involved.  You may be impacted by the 
default 15 second bridge forwarding timeout.

By by default, it takes the bridge 15 seconds to accept packets on a 
newly joined interface.  This is a spanning tree issue.

You can check the value with brctl:

/usr/sbin/brctl  showstp ${DEVICE}


You set it with brctl like this:

/usr/sbin/brctl setfd ${DEVICE} ${DELAY}


However, it is much more convenient to set it on the ifcfg-brN

Simply add a DELAY parameter to the br0 definition.

You can see the example on the libvirt wiki page, although they do not 
explain it there.

http://wiki.libvirt.org/page/Networking

It took us a while to find it, because the VMs start quicker on RHEL-6.  
On RHEL-5 based KVMs, we always noticed how long the PXE boot took, but 
it did eventually start.  That is because it took 5 seconds longer to 
get to that point.

When we did find it, and correct it on both new and old servers, the 
RHEL-5 based VMs PXE booted immediately as well.


Good Luck!

_______________________________________________
rhelv6-list mailing list
rhelv6-list at redhat.com
https://www.redhat.com/mailman/listinfo/rhelv6-list