ccfcfc
asked on
Sendmail error "lost input channel from" using Sendmail version 8.13.1
Using Sendmail version 8.13.1 from a Linux box.
Seem to be getting " lost input channel " and a "User Unknown " almost immediate. Can send email to every other external domains i.e @hotmail.com or @gmail.com but if the destination domain is the company's domain then it gets rejected.
This sendmail in on an internal network and allows mail out /in and due to other mail going out we know the firewall is not to blame for rejection. The mails being sent go via the correct SMTP server as seen in the mail log.
I have added a RELAY to this domain and even SPAM CONTROL for that Domain to ALLOW . But every time we get "lost input channel" and almost immediately the <username@companyname.com user Unknown then the "lost input channel".
It is really weird as to why we are unable to send to every other domain than our own.
Appreciate as to why .
Seem to be getting " lost input channel " and a "User Unknown " almost immediate. Can send email to every other external domains i.e @hotmail.com or @gmail.com but if the destination domain is the company's domain then it gets rejected.
This sendmail in on an internal network and allows mail out /in and due to other mail going out we know the firewall is not to blame for rejection. The mails being sent go via the correct SMTP server as seen in the mail log.
I have added a RELAY to this domain and even SPAM CONTROL for that Domain to ALLOW . But every time we get "lost input channel" and almost immediately the <username@companyname.com user Unknown then the "lost input channel".
It is really weird as to why we are unable to send to every other domain than our own.
Appreciate as to why .
ASKER
Emails go out as you can see below so I know SMTP traffic is working. The MX record works and is valid as it is used for company emails in general. SO I know the Domain/MX is correct
See extract from maillog , hope this helps
This one worked - see destination address :-
from=<Info@Intamac.com>, size=639, class=0, nrcpts=1, msgid=<200711281147.lASBlH Ew011401@m ail01.dev> , proto=ESMTP, daemon=Daemon0, relay=inta-dev-iis01.dev.i ntamac.com [192.168.31.40]
Nov 28 11:47:18 mail01 sendmail[11403]: lASBlHEw011401: to=<m.farrington1980@gmail .com>, delay=00:00:01, xdelay=00:00:01, mailer=relay, pri=120639, relay=smtp.bt.net. [194.72.6.62], dsn=2.0.0, stat=Sent (OK id=1IxLLu-0005Is-M1)
This one not - same application creating email . For some reason it tried to use different SMTP agent
<mFarrington@intamac.com>. .. User unknown
Nov 28 12:11:05 mail01 sendmail[11532]: lASCB53r011532: lost input channel from inta-dev-iis01.dev.intamac .com [192.168.31.40] to Daemon0 after rcpt
Nov 28 12:11:05 mail01 sendmail[11532]: lASCB53r011532: from=<Info@Intamac.com>, size=0, class=0, nrcpts=0, proto=ESMTP, daemon=Daemon0, relay=inta-dev-iis01.dev.i ntamac.com [192.168.31.40]
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: <mFarrington@intamac.com>. .. User unknown
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: lost input channel from inta-dev-iis01.dev.intamac .com [192.168.31.40] to Daemon0 after rcpt
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: from=<Info@Intamac.com>, size=0, class=0, nrcpts=0, proto=ESMTP, daemon=Daemon0, relay=inta-dev-iis01.dev.i ntamac.com [192.168.31.40]
See extract from maillog , hope this helps
This one worked - see destination address :-
from=<Info@Intamac.com>, size=639, class=0, nrcpts=1, msgid=<200711281147.lASBlH
Nov 28 11:47:18 mail01 sendmail[11403]: lASBlHEw011401: to=<m.farrington1980@gmail
This one not - same application creating email . For some reason it tried to use different SMTP agent
<mFarrington@intamac.com>.
Nov 28 12:11:05 mail01 sendmail[11532]: lASCB53r011532: lost input channel from inta-dev-iis01.dev.intamac
Nov 28 12:11:05 mail01 sendmail[11532]: lASCB53r011532: from=<Info@Intamac.com>, size=0, class=0, nrcpts=0, proto=ESMTP, daemon=Daemon0, relay=inta-dev-iis01.dev.i
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: <mFarrington@intamac.com>.
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: lost input channel from inta-dev-iis01.dev.intamac
Nov 28 12:12:10 mail01 sendmail[11550]: lASCC9gd011550: from=<Info@Intamac.com>, size=0, class=0, nrcpts=0, proto=ESMTP, daemon=Daemon0, relay=inta-dev-iis01.dev.i
Does mail01 /etc/mail/local-host-names have 'intamac.com'?
If so, are the intamac.com mailboxes on mail01?
Is it possible that mail01 is trying to deliver the mail locally and shouldn't?
If so, are the intamac.com mailboxes on mail01?
Is it possible that mail01 is trying to deliver the mail locally and shouldn't?
ASKER
Yes /etc /mail/local-host-names does have intamac.com in it.
No the mail boxes are not on mail01 but on an external domain used by the company.
Well I kinda guessed that mail01 was trying to do that but and should not as the username for the intamac.com domain are external. Whats is confusing if we sent to any other domain they get sent as expected.
Like I said I have added SPAN CONTROL entry and a relay entry to allow to go to intamac.com . It is as you said, it is trying to deliver to the local mail01 server of which the require dmailbox does not exist on.
I guess I need to force mail out if it says <usrename>@intamac.com . Not entirely sure as to why is this not happening already.
No the mail boxes are not on mail01 but on an external domain used by the company.
Well I kinda guessed that mail01 was trying to do that but and should not as the username for the intamac.com domain are external. Whats is confusing if we sent to any other domain they get sent as expected.
Like I said I have added SPAN CONTROL entry and a relay entry to allow to go to intamac.com . It is as you said, it is trying to deliver to the local mail01 server of which the require dmailbox does not exist on.
I guess I need to force mail out if it says <usrename>@intamac.com . Not entirely sure as to why is this not happening already.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
Thanks, removing the local domain fixed the problem. Kinda just needed confirmation of what I was thinking.
Does your sendmail box think that your domain is local to that machine (/etc/mail/local-host-name
Is the MX record for your domain pointing to the correct fully qualified domain name?
Could you post the /var/log/maillog snippet?