Hello. I was very surprised yesterday. I was watching mpeg4 video on
my linux system over the 10Mbps local net. Video was mounted as smb
share and was played well.
My machine is a local IRC server, and there were 5-6 users on the
IRC. Then, the movie stopped occasionally. I was not surprised, may be
smb server with mpeg4 was down. But I found, that users on the IRC cry
about network bugs and so..
I start ping to the local gateway:
PING 10.6.0.10 (10.6.0.10) from 10.6.6.6 : 56(84) bytes of data.
64 bytes from 10.6.0.10: icmp_seq=0 ttl=64 time=1022.5 ms
64 bytes from 10.6.0.10: icmp_seq=1 ttl=64 time=533.7 ms
64 bytes from 10.6.0.10: icmp_seq=2 ttl=64 time=434.5 ms
64 bytes from 10.6.0.10: icmp_seq=3 ttl=64 time=3002.5 ms
...
And this is over 10MBps. I was very surprised. The next command was:
ifconfig - no hardware bugs shows to me. Then I start tcpdump and found
an ftp traffic from my machine. Then I start 'iptraf' and found, that
user from my segment downloading a large file from me by ftp. iptraf
shows a 1040 Kb/sec traffic.
So, when he stopped the downloading, ping go to normal state and my
movie continue to play.
PING 10.6.0.10 (10.6.0.10) from 10.6.6.6 : 56(84) bytes of data.
64 bytes from 10.6.0.10: icmp_seq=0 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=1 ttl=64 time=0.7 ms
64 bytes from 10.6.0.10: icmp_seq=2 ttl=64 time=0.5 ms
Then I repeat an expirement, in one console I start ping, in the
another, I start uploading file from my machine to another in my segment
by the ftp.
PING 10.6.0.10 (10.6.0.10) from 10.6.6.6 : 56(84) bytes of data.
64 bytes from 10.6.0.10: icmp_seq=0 ttl=64 time=1.0 ms
64 bytes from 10.6.0.10: icmp_seq=1 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=2 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=3 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=4 ttl=64 time=0.5 ms
------- uploading started:
64 bytes from 10.6.0.10: icmp_seq=5 ttl=64 time=324.0 ms
64 bytes from 10.6.0.10: icmp_seq=6 ttl=64 time=523.3 ms
64 bytes from 10.6.0.10: icmp_seq=7 ttl=64 time=1002.5 ms
64 bytes from 10.6.0.10: icmp_seq=8 ttl=64 time=3424.5 ms
64 bytes from 10.6.0.10: icmp_seq=9 ttl=64 time=1231.3 ms
So, it is not only ping. Any network traffic was 'died' when starting
ftp upload/download in my segment. If try to download/upload to another
segment ping is:
64 bytes from 10.6.0.10: icmp_seq=149 ttl=64 time=1.9 ms
64 bytes from 10.6.0.10: icmp_seq=150 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=151 ttl=64 time=0.5 ms
64 bytes from 10.6.0.10: icmp_seq=152 ttl=64 time=1.8 ms
---------- uploading started:
64 bytes from 10.6.0.10: icmp_seq=153 ttl=64 time=2.4 ms
64 bytes from 10.6.0.10: icmp_seq=154 ttl=64 time=9.8 ms
64 bytes from 10.6.0.10: icmp_seq=155 ttl=64 time=11.6 ms
64 bytes from 10.6.0.10: icmp_seq=156 ttl=64 time=9.0 ms
64 bytes from 10.6.0.10: icmp_seq=157 ttl=64 time=9.4 ms
64 bytes from 10.6.0.10: icmp_seq=158 ttl=64 time=12.6 ms
Does the same problem repeats on your systems?
-
: send the line "unsubscribe linux-net" in
the body of a message to majordomo@vger.kernel.org
