This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-API] Re: openvswitch problem with XCP 1.0 RC3

To: Adrian Costin <adrian@xxxxxxxxxxx>
Subject: Re: [Xen-API] Re: openvswitch problem with XCP 1.0 RC3
From: Christian Fischer <christian.fischer@xxxxxxxxxxxxxxxxxxx>
Date: Mon, 28 Feb 2011 09:52:09 +0100
Cc: xen-api@xxxxxxxxxxxxxxxxxxx
Delivery-date: Mon, 28 Feb 2011 00:52:26 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <87pqqfidg1.fsf@xxxxxxxxxxxxxxxx>
List-help: <mailto:xen-api-request@lists.xensource.com?subject=help>
List-id: Discussion of API issues surrounding Xen <xen-api.lists.xensource.com>
List-post: <mailto:xen-api@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-api>, <mailto:xen-api-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-api>, <mailto:xen-api-request@lists.xensource.com?subject=unsubscribe>
Organization: EasternGraphics GmbH
References: <EBDA14578E54D94191DD90533575FDFBF5EB3816@xxxxxxxxxxxxxxxxxxxxx> <87pqqfidg1.fsf@xxxxxxxxxxxxxxxx>
Sender: xen-api-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: KMail/1.13.5 (Linux/2.6.29-gentoo-r5; KDE/4.4.5; x86_64; ; )
On Saturday 26 February 2011 06:15:26 Ben Pfaff wrote:
> Adrian Costin <adrian@xxxxxxxxxxx> writes:
> > When using the openvswitch backend (which is the default) on XCP 1.0 RC3
> > network traffic for VLANs doesn’t reach the virtual machine.
> > 
> > Switching to bridge works without problems.
> Some kernel network drivers, including igb, have VLAN-related
> bugs that affect OVS but not the Linux bridge.  At Nicira we're
> working on a workaround for this.  It will likely be ready next
> week.

A simple ping should always work.

There are two problems with igb/e1000e/e1000.

You must patch the drivers to avoid chksum errors like
e1000_tx_csum: checksum_partial proto=81!

You must set the external port MTU to 1518 or VMs MTU to 1492

We have no problems at all since we've applied the patches and adjusted the 
external port MTU settings. All Networks are VLAN tagged, including the 
Control network. All externel switch ports are vlan trunks tagged with the 
appropriate VLAN IDs.

xen-api mailing list