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

RE: [Xen-users] XCP ethernet jumbo frames????

  • To: Gémes Géza <geza@xxxxxxxxxxx>, <xen-users@xxxxxxxxxxxxxxxxxxx>
  • From: "James Harper" <james.harper@xxxxxxxxxxxxxxxx>
  • Date: Tue, 21 Sep 2010 08:27:45 +1000
  • Cc:
  • Delivery-date: Mon, 20 Sep 2010 15:29:04 -0700
  • List-id: Xen user discussion <xen-users.lists.xensource.com>
  • Thread-index: ActY98V9vUpZMof3RGSypvtONh36/QAGwxHg
  • Thread-topic: [Xen-users] XCP ethernet jumbo frames????

> Hi,
> I have an XCP 0.5 box running a few DomUs some of the DomU's data is
> stored at (clustered lvm volumes on) a Coraid etherdrive box which is
> connected via an Ethernet cable to a dedicated card of the XCP box, for
> which in the network config I' ve specified: MTU=9344: xe
> network-param-list uuid=.... gives:
> MTU ( RW): 9344
> on the DomUs if I leave the MTU at 1500 (default settings) everything
> works (slowly, but works). If I increase the MTU the aoe device appear
> as expected, but the data on it becomes inaccessible, even just trying
> to activate the lvm volume groups gives a bunch of IO errors at sector
> 0, and so on.
> Please tell me what I'm doing wrong. Are jumbo frames expected et all to
> work?

Does your network adapter really support jumbo frames?
Is the network adapter on a bridge? (I'm unsure if Dom0 does AoE or if the 
DomU's do)
Are all the bridge interfaces set to 9344 MTU?
Could tcp offload settings break any of this? (unlikely - AOE is not TCP)

tcpdump might tell you what's going on.


Xen-users mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.