On Jun 26, 2017, at 5:57 PM, Zheng, Ruoqin <zhengrq.f...@cn.fujitsu.com> wrote:
> Hi Guy > Thank you for your answer, and here is the output with -d: > > root # ping -q -c 50 -I eth0.100 192.168.255.2& > [2] 1208 > root # tcpdump -d -i eth0 ether broadcastPING 192.168.255.2 > (192.168.255.2) from 192.168.255.1 eth0.1 > > (000) ld [2] > (001) jeq #0xffffffff jt 2 jf 5 > (002) ldh [0] > (003) jeq #0xffff jt 4 jf 5 > (004) ret #262144 > (005) ret #0 > > root # tcpdump -n -d -i eth0 ether broadcast > (000) ld [2] > (001) jeq #0xffffffff jt 2 jf 5 > (002) ldh [0] > (003) jeq #0xffff jt 4 jf 5 > (004) ret #262144 > (005) ret #0 > root # > > You can see they are same. As I expected. So what happens if you run tcpdump, without -n, do the ping, and wait a long time to see whether tcpdump eventually prints something? It might be taking a long time to try to resolve a host name, for some reason or another, so that it can't print anything until that attempt either succeeds or times out. _______________________________________________ tcpdump-workers mailing list tcpdump-workers@lists.tcpdump.org https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers