cvs commit: src/lib/libfetch http.c
Pawel Worach
pawel.worach at telia.com
Fri Feb 18 23:02:42 GMT 2005
Colin Percival wrote:
> Pawel Worach wrote:
>
>> This seems to have broken fetch though a http proxy (in this case
>> squid on Solaris).
>>
>> tcpdump on proxy box:
>
> > [snip]
>
> The tcpdump shown looks entirely correct, but it's hard to know without
> seeing the actual data... could you re-run this with tcpdump capturing
> the packet payloads as well?
>
> Colin Percival
>
Sure.
# /opt/sfw/sbin/tcpdump -vvv -tt -xX -s0 port 8080 and host client
1108767288.666846 client.53107 > proxy.8080: S [tcp sum ok]
341123682:341123682(0) win 65535 <mss 1460,nop,nop,sackOK,nop,wscale
1,nop,nop,timestamp 205501788 0> (DF) (ttl 64, id 29183, len 64)
0x0000 4500 0040 71ff 4000 4006 710d 0aa3 a133 E.. at q.@. at .q....3
0x0010 0aa3 a132 cf73 1f90 1455 2262 0000 0000 ...2.s...U"b....
0x0020 b002 ffff f6fd 0000 0204 05b4 0101 0402 ................
0x0030 0103 0301 0101 080a 0c3f b55c 0000 0000 .........?.\....
1108767288.666875 proxy.8080 > client.53107: S [tcp sum ok]
1666873470:1666873470(0) ack 341123683 win 49232 <nop,nop,timestamp 2616716056
205501788,mss 1460,nop,wscale 0,nop,nop,sackOK> (DF) (ttl 64, id 56931, len 64)
0x0000 4500 0040 de63 4000 4006 04a9 0aa3 a132 E.. at .c@. at ......2
0x0010 0aa3 a133 1f90 cf73 635a 787e 1455 2263 ...3...scZx~.U"c
0x0020 b012 c050 d3b3 0000 0101 080a 9bf7 eb18 ...P............
0x0030 0c3f b55c 0204 05b4 0103 0300 0101 0402 .?.\............
1108767288.667069 client.53107 > proxy.8080: . [tcp sum ok] 1:1(0) ack 1 win
33304 <nop,nop,timestamp 205501788 2616716056> (DF) (ttl 64, id 29185, len 52)
0x0000 4500 0034 7201 4000 4006 7117 0aa3 a133 E..4r. at .@.q....3
0x0010 0aa3 a132 cf73 1f90 1455 2263 635a 787f ...2.s...U"ccZx.
0x0020 8010 8218 52b7 0000 0101 080a 0c3f b55c ....R........?.\
0x0030 9bf7 eb18 ....
1108767288.667432 client.53107 > proxy.8080: FP [tcp sum ok] 1:113(112) ack 1
win 33304 <nop,nop,timestamp 205501788 2616716056> (DF) (ttl 64, id 29186, len 164)
0x0000 4500 00a4 7202 4000 4006 70a6 0aa3 a133 E...r. at .@.p....3
0x0010 0aa3 a132 cf73 1f90 1455 2263 635a 787f ...2.s...U"ccZx.
0x0020 8019 8218 a3b9 0000 0101 080a 0c3f b55c .............?.\
0x0030 9bf7 eb18 4745 5420 6874 7470 3a2f 2f6e ....GET.http://n
0x0040 6577 732e 6262 632e 636f 2e75 6b2f 2048 ews.bbc.co.uk/.H
0x0050 5454 502f 312e 310d 0a48 6f73 743a 206e TTP/1.1..Host:.n
0x0060 6577 732e 6262 632e 636f 2e75 6b0d 0a55 ews.bbc.co.uk..U
0x0070 7365 722d 4167 656e 743a 2066 6574 6368 ser-Agent:.fetch
0x0080 206c 6962 6665 7463 682f 322e 300d 0a43 .libfetch/2.0..C
0x0090 6f6e 6e65 6374 696f 6e3a 2063 6c6f 7365 onnection:.close
0x00a0 0d0a 0d0a ....
1108767288.667478 proxy.8080 > client.53107: . [tcp sum ok] 1:1(0) ack 114 win
49120 <nop,nop,timestamp 2616716056 205501788> (DF) (ttl 64, id 56932, len 52)
0x0000 4500 0034 de64 4000 4006 04b4 0aa3 a132 E..4.d at .@......2
0x0010 0aa3 a133 1f90 cf73 635a 787f 1455 22d4 ...3...scZx..U".
0x0020 8010 bfe0 147e 0000 0101 080a 9bf7 eb18 .....~..........
0x0030 0c3f b55c .?.\
1108767288.669709 proxy.8080 > client.53107: F [tcp sum ok] 1:1(0) ack 114 win
49232 <nop,nop,timestamp 2616716056 205501788> (DF) (ttl 64, id 56934, len 52)
0x0000 4500 0034 de66 4000 4006 04b2 0aa3 a132 E..4.f at .@......2
0x0010 0aa3 a133 1f90 cf73 635a 787f 1455 22d4 ...3...scZx..U".
0x0020 8011 c050 140d 0000 0101 080a 9bf7 eb18 ...P............
0x0030 0c3f b55c .?.\
1108767288.669968 client.53107 > proxy.8080: . [tcp sum ok] 114:114(0) ack 2 win
33303 <nop,nop,timestamp 205501791 2616716056> (DF) (ttl 64, id 29188, len 52)
0x0000 4500 0034 7204 4000 4006 7114 0aa3 a133 E..4r. at .@.q....3
0x0010 0aa3 a132 cf73 1f90 1455 22d4 635a 7880 ...2.s...U".cZx.
0x0020 8010 8217 5243 0000 0101 080a 0c3f b55f ....RC.......?._
0x0030 9bf7 eb18 ....
Solaris box is: SunOS proxy 5.9 Generic_112233-12 sun4u sparc SUNW,Ultra-4
FreeBSD box is: FreeBSD client 6.0-CURRENT FreeBSD 6.0-CURRENT #0: Wed Feb 16
14:39:08 CET 2005
Squid is: Squid Cache: Version 2.5.STABLE6-20041001
Both machines are connected to the same switch and are on the same subnet.
--
Pawel
More information about the cvs-src
mailing list