Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 16 Oct 1998 18:07:30 -0600 (MDT)
From:      qed <baram@mscd.edu>
To:        freebsd-questions@FreeBSD.ORG
Subject:   inn1.7.2 problems:(
Message-ID:  <Pine.OSF.3.95.981016173805.18584A-100000@clem.mscd.edu>

next in thread | raw e-mail | index | archive | help
Hi all,
I have been experiencing annoying problems with inn1.7.2 on
freebsd-2.2.7-release. My feeder(what's defined in
$INN/etc/hosts.nntp && $INN/etc/nntpsend.ctl) happends to be
news.uu.net. I regurarly run nntpsend from crontab to feed up whatever was
batched up in the prior 10 minutes. Now for the problem:(
If i have an article that is larger than 300 bytes in size(approximate),
innxmit(which runs every 10 min) timeouts and doesnt transfer any of the
articles. Just for the kicks, i tried running nntpsend with -d/-D option,
and noticed that innxmit transfers the articles ok, upto the last dot,
after which it waits for responce. Since it doesnt get any reply back
within the allocated time, i get that timeout error. So i start up ktrace:
# ktrace -tcnisuw -i /usr/news/bin/nntpsend -d
get a little kdump.out file..with the relavant output after the articles
and the last ".\r" below:
        .\r
       "
 12341 innxmit  RET   write 1431/0x597
 12341 innxmit  CALL  select(0x6,0xefbfb854,0,0,0xefbfb84c)
 12341 innxmit  CSW  stop kernel
 12341 innxmit  CSW  resume kernel
 12341 innxmit  RET   select 1
 12341 innxmit  CALL  read(0x5,0x96e8,0x400)
 12341 innxmit  RET   read -1 errno 60 Operation timed out
 12341 innxmit  CALL  write(0x2,0xefbfb144,0x27)
 12341 innxmit  GIO   fd 2 wrote 39 bytes
       "No reply to check, Operation timed out
       "
 12341 innxmit  RET   write 39/0x27
 12341 innxmit  CALL  write(0x2,0xefbfb344,0x22)
 12341 innxmit  GIO   fd 2 wrote 34 bytes
       "Rewriting batch file and exiting.
       "
 12341 innxmit  RET   write 34/0x22

Ok..so i i decide to run tcpdump, since this obviosly looks like the
article got written to a socket . By this time i start having suspicions
that this might be a kernel(or driver) issue, because telnet'ing directly
to nntp-in.uu.net 119 and posting a large article hangs as well. So i run
tcpdump..and interestingly enough, only part of the last article gets
transfered, so the remote end never sees the last ".\r" and doesnt
acknoledge(thus the timeout). I have verified that this is not a packet
filter issue by posting a large enough article from machine(bsd/os) inside
and outside the filter. I am using an ep0 driver w/ 3c5x9 which is
detected fine, and works fine(news feed is quite large and never
any problems with it). This is probabaly not a bug in inn either, since
posting directly via tlnet doesnt work. Ok, sorry for such a long post. If
you need any more info, ill be glad to provide it.
Thank you,
	Alex


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.OSF.3.95.981016173805.18584A-100000>