On 03/04/2010 11:47 PM, Søren Ragsdale wrote: > On Mar 4, 2010, at 12:05 AM, Min Miles Xu wrote: > > >> It does look strange to me. I expect to get ICMP packets or ARP packets. But >> I notice all the packets the Solaris machine got are TCP packets. The sender >> is 10.0.1.3. Is it the machine that was echoing the ping packets? Are the >> two machines back-to-back connected?Here is the parsing result of one packet. >> >> >> >> desc mac addr: 00 25 90 01 33 08 >> >> src mac addr: 00 1d 7d e6 af db >> >> type: 08 00 (IPv4) >> >> Version: 4 >> >> IHL: 5 >> >> DS: 00 >> >> Total Length: 01 06 >> >> ID: 28 f2 >> >> Flag & Offset: 40 00 >> >> TTL: 80 >> >> Protocal: 06 (TCP) >> >> Header Checksum: ba fb >> >> src IP addr: 0a 00 01 03 >> >> desc IP addr: 0a 00 01 02 >> >> c0 e5 01 bd 1b 20 a2 35 0b 62 86 45 50 18 >> >> 3f e3 a4 02 00 00 00 00 00 da ff 53 ff >> > There are at least three machines on this network: > > 10.0.1.2 (00:25:90:01:33:08): opensolaris > 10.0.1.3 (00:1d:7d:e6:af:db): windows > 10.0.1.4 (00:25:4b:9c:c8:b6): OSX > > All three machines are connected to a Netgear GS105 switch, which is > connected to an Apple AirPort Extreme router: > > opensolaris---\ > windows-------GS105-----AirPort > osx ----------/ > > > I've always sent pings from 10.0.1.2 to 10.0.1.4, so that the switch would > have no excuses to drop any pings. 10.0.1.3 might be generating some traffic > and is probably trying to restore its lost CIFS share, but its traffic is > irrelevant to our tests. > > If you'd like, I can try to re-run the test after unplugging 10.0.1.3 to > eliminate its potential interference with our problem. >
The interface is good to receive the packets since it got packets from the Windows machine. But the ping packets was dropped somewhere. If it's convenient for you, could you simplify the configurations to have two machines back-to-back connected. _______________________________________________ networking-discuss mailing list networking-discuss@opensolaris.org
