lkml.org 
[lkml]   [2000]   [Mar]   [24]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectLinux + Apache + Virtualhost problem
My apologies if this is not the proper forum for this report.

Server: Slackware 7, Linux kernel 2.2.14, Apache 1.3.11, RealTek RTL-8029
NIC (PCI NE2000 clone, using ne2k-pci.c compiled into the kernel)

Problem: Apache configured using IP-based virtual hosting. IP is active
and pingable. Apache refuses to respond to SYN packets sent from client
to a virtual host served by Apache. The SYN's come in but the ACK is not
sent, at least not right away. After a delay (1 minute or so) either the
client times out or the connection finally goes through. If the client's
browser does not timeout, and the connection finally goes through, there
are no more noticable delays until the client disconnects or tries again
later.

Client sits in SYN_SENT, server in SYN_RECV. If, while the client's
browser is trying to connect, the client telnets to the server, both the
telnet and web connections immediately go through! All previously
explained delay goes away. If the client disconnects from ISP, or tries
again later, problem reoccurs.

The problem occurs on every client machine I have tried: Linux, Windows,
Netscape, Lynx...

Another server on the same network with a _very_ similar Apache / network
configuration does not suffer this problem. It is using an Intel eepro
NIC. Otherwise configuration is the same.

Restarting Apache / rebooting server seems to make the problem go away for
a short period of time.

You can see this behavior for yourself: www.wolfclan2000.com
(209.96.226.205)

Thank you for any help you can offer. I have searched all archives I
could find and spoken with other Slackware folks about it and no one has
found an answer. Below you will find a TCP dump of the server machine
showing the many SYN packets that had to come through before an ACK was
finally sent. Netscape usually times out before this happens.

Please CC any replies directly to me.

Best Regards,

Thomas Duffey
cortex@execpc.com

##[ tcpdump ]##
blaze:~# tcpdump -nvv '(host 209.96.226.205 and 4.19.114.82) and tcp and port 80 ^M'
tcpdump: listening on eth0
10:21:32.589816 4.19.114.82.12964 > 209.96.226.205.80: S 3653171195:3653171195(0) win 32120 <mss 1460,sackOK,timestamp 5649041[|tcp]> (DF) (ttl 52, id 39732)
10:21:35.547105 4.19.114.82.12964 > 209.96.226.205.80: S 3653171195:3653171195(0) win 32120 <mss 1460,sackOK,timestamp 5649341[|tcp]> (DF) (ttl 52, id 39752)
10:21:41.547345 4.19.114.82.12964 > 209.96.226.205.80: S 3653171195:3653171195(0) win 32120 <mss 1460,sackOK,timestamp 5649941[|tcp]> (DF) (ttl 52, id 39790)
10:21:53.579271 4.19.114.82.12964 > 209.96.226.205.80: S 3653171195:3653171195(0) win 32120 <mss 1460,sackOK,timestamp 5651141[|tcp]> (DF) (ttl 52, id 39835)
10:22:17.567454 4.19.114.82.12964 > 209.96.226.205.80: S 3653171195:3653171195(0) win 32120 <mss 1460,sackOK,timestamp 5653541[|tcp]> (DF) (ttl 52, id 39843)
10:22:17.567504 209.96.226.205.80 > 4.19.114.82.12964: S 3311182500:3311182500(0) ack 3653171196 win 32120 <mss 1460,sackOK,timestamp 34025263[|tcp]> (DF) (ttl 64, id 9823)
10:22:17.725096 4.19.114.82.12964 > 209.96.226.205.80: . ack 1 win 32120 <nop,nop,timestamp 5653558 34025263> (DF) (ttl 52, id 39844)
10:22:17.775157 4.19.114.82.12964 > 209.96.226.205.80: P 1:286(285) ack 1 win 32120 <nop,nop,timestamp 5653558 34025263> (DF) (ttl 52, id 39845)
10:22:17.775228 209.96.226.205.80 > 4.19.114.82.12964: . ack 286 win 32120 <nop,nop,timestamp 34025284 5653558> (DF) (ttl 64, id 9827)
10:22:17.776592 209.96.226.205.80 > 4.19.114.82.12964: P 1:488(487) ack 286 win 32120 <nop,nop,timestamp 34025284 5653558> (DF) (ttl 64, id 9828)
10:22:17.946502 4.19.114.82.12964 > 209.96.226.205.80: . ack 488 win 32120 <nop,nop,timestamp 5653581 34025284> (DF) (ttl 52, id 39849)

11 packets received by filter
0 packets dropped by kernel

##[ netstat ]##
tcp 0 0 www.wolfclan2000.co:www ip082.wmdesign.ne:12964 SYN_RECV


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:57    [W:0.023 / U:1.120 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site