ltspbr0 & ethx packet descripincy

[Barry R Cisna]

Hello All,

CentOS - 6.3 32-bit
SL - 6.0
ltsp 5.0

I was just looking at Wireshark,,,for a while today on one of the
k12linux / ltsp5 servers. With Wireshark open and not doing any actual
'capturing', when I click on the far left icon on the toolbar of
Wireshark,which shows you the available interface capture devices.
You can see the packets going in & out of each interface.

When looking at ltspbr0 & eth0 ( which are the bridged interfaces),,,the
ltspbr0 interface shows several less incoming and outgoing packets,,,in
just a short amount of time.
As dumb as I am I would naturally think the two bridged interfaces
should be showing identical packet count.

Has anyone else on board here ever seen this,,or maybe have never
stumbled across this ? I never really calculated in a percentage,but I
can not help but think,this is hampering network throughput to some
degree or another.
I do not want to spend a bunch of time,,trying to over anaylize this but
just was curious if anyone else may be seeing the same behaviour?

Thanks,
Barry

[Barry R Cisna]

Hello All,

Wanted to add this for completeness.
I do not understand why so much difference between ltspbr0 & eth0, the
bridged interfaces.
Seems these two should be theoritically identical?
I guess no symmetrical going on here. Never noticed this before,I guess?
rx & tx between the two seems way out of whack?

ifconfig,,,

eth0 Link encap:Ethernet HWaddr 00:00:00:00:59:11
inet6 addr: 0000::000:4cff:fe20:5911/64 Scope:Link
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:79411529 errors:0 dropped:0 overruns:0 frame:0
TX packets:109054017 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:775690448 (739.7 MiB) TX bytes:1766710101 (1.6 GiB)
Interrupt:26 Base address:0xe000

ltspbr0 Link encap:Ethernet HWaddr 00:00:00:00:59:11
inet addr:172.31.100.254 Bcast:172.31.100.255
Mask:255.255.255.0
inet6 addr: 0000::000:4cff:fe20:5911/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:34405727 errors:0 dropped:0 overruns:0 frame:0
TX packets:47671964 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:2480244830 (2.3 GiB) TX bytes:546859657 (521.5 MiB)

Someone smarter than me,I am sure knows why this is..

Thanks,
Barry

[Carl Keil]

Could it be that txqueuelen setting on eth0? There's a difference
there. I tried googling this, but haven't found out exactly what that
is or does though. I could see forcing a certain queue size adding
packets to what eventually gets sent, if there is also some minimum time
interval to flushing the queue or something. But I'm certainly not
smarter than you.

ck

snip

Post by Barry R Cisna
eth0 Link encap:Ethernet HWaddr 00:00:00:00:59:11
inet6 addr: 0000::000:4cff:fe20:5911/64 Scope:Link
UP BROADCAST RUNNING PROMISC MULTICAST MTU:1500 Metric:1
RX packets:79411529 errors:0 dropped:0 overruns:0 frame:0
TX packets:109054017 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:775690448 (739.7 MiB) TX bytes:1766710101 (1.6 GiB)
Interrupt:26 Base address:0xe000

/snip

Post by Barry R Cisna
Thanks,
Barry