Sendmail bug
Web Dove
dove at MIT-BUGS-BUNNY.ARPA
Fri Feb 22 15:06:35 AEST 1985
We use sendmail in a mode that makes direct connections to other sites.
I have occasionally seen it return error "Bad File Number" while talking to
some hosts.
Apparently everything is fine but the last reply. The timeout is set
to 2hrs, so that is not the problem. I tried
"/usr/src/usr.lib/sendmail/aux/mconnect maryland" and did the transfer by
hand and it worked. Does this sound familiar to anyone?
The following transcript fragment comes from sendmail -d -v -q:
SENDALL: mode i, sendqueue:
364c4=liz at maryland: mailer 4 (tcp), host `maryland', user `liz at maryland'
next=0, flags=10, alias 0
home="", fullname=""
recipient: 13844=eem: mailer 1 (local), host `', user `eem'
next=0, flags=1, alias 0
home="/syla/dspg/eem", fullname=""
--deliver, mailer=4, host=`maryland', first user=`liz at maryland'
remotename(eem)
remotename => `eem at MIT-BUGS-BUNNY.ARPA'
send to 364c4=liz at maryland: mailer 4 (tcp), host `maryland', user `liz at maryland'
next=13844, flags=10, alias 0
home="", fullname=""
liz at maryland... Connecting to maryland.tcp...
openmailer: "IPC" "maryland"
makeconnection (maryland)
makeconnection: 9
reply
220 maryland.ARPA Sendmail 4.12/4.7 ready at Thu, 21 Feb 85 11:50:41 est
>>> HELO MIT-BUGS-BUNNY.ARPA
reply
250 maryland.ARPA Hello MIT-BUGS-BUNNY.ARPA, pleased to meet you
>>> MAIL From:<eem at MIT-BUGS-BUNNY.ARPA>
reply
250 <eem at MIT-BUGS-BUNNY.ARPA>... Sender ok
remotename(liz at maryland)
remotename => `liz at maryland'
>>> RCPT To:<liz at maryland>
reply
250 <liz at maryland>... Recipient ok
>>> DATA
reply
354 Enter mail, end with "." on a line by itself
remotename(Evangelos Milios <eem>)
crackaddr(Evangelos Milios <eem>)
crackaddr=>`Evangelos Milios <$g>'
remotename => `Evangelos Milios <eem at MIT-BUGS-BUNNY.ARPA>'
remotename(liz at maryland)
crackaddr(liz at maryland)
crackaddr=>`$g'
remotename => `liz at maryland'
>>> .
reply
liz at maryland... reply: read error
liz at maryland... Deferred: Bad file number
====finis: stat 0 e_flags 2
More information about the Comp.unix.wizards
mailing list