From: David Coulson <david@da...> - 2002-09-15 09:10:07
I switched my UMLs over to use a kernel bridge, rather than uml_switch,
so I've got a bunch of tap devices sitting on the host for connectivity
to the UML. However, there seems to be some discrepency between the TX
bytes on the tap device, and the RX bytes on the UML.
uml-web0 Link encap:Ethernet HWaddr 00:FF:6C:6D:70:36
inet6 addr: fe80::2ff:6cff:fe6d:7036/10 Scope:Link
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:1411334 errors:0 dropped:0 overruns:0 frame:0
TX packets:1361146 errors:0 dropped:1 overruns:0 carrier:0
RX bytes:262129828 (249.9 MiB) TX bytes:246860369 (235.4 MiB)
eth0 Link encap:Ethernet HWaddr FE:FD:0A:01:07:05
inet addr:10.1.7.5 Bcast:10.255.255.255 Mask:255.255.255.0
inet6 addr: fe80::fcfd:aff:fe01:705/10 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1361150 errors:0 dropped:0 overruns:0 frame:0
TX packets:1411336 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:227804597 (217.2 MiB) TX bytes:262130024 (249.9 MiB)
The number of packets is near enough identical, but there's a
significant difference (~7%) between bandwidth usage of the tap device,
and the interface on the UML side. Does anyone have an idea what
could/does account for this? I'm intending setting up SNMP bandwidth
accounting, but if the values of the host tap device don't match that of
the UML, then people are going to complain. The RX on the host is the
same as the TX on the UML, so I'm sort of confused as to why that's
right, when bandwidth usage of packets to the UML isn't.
All my other UMLs on the host exhibit the same effect.
David Coulson http://davidcoulson.net/
Get latest updates about Open Source Projects, Conferences and News.