sendmail config for workflow mailer

[Meakin Singh]

hi,

currently iam configuring workflow mail for EBS 11.5.10.2, but i ned to configure senmail for that purpose , my OS is HP UX 11.11

we do have a seperate mail server in our organisation , can i send mail directly from my Unix server to mail clients configured with the mail server which i suppose is configured with some version of linux, we use outlook express as a mail client on windows platform

[TTr]

It is very common to have a single mailbox so that your users don't have to login to more than one email system to get their email. So unless you are overwhelmingly asked otherwise by your users or management, the outlook mailbox should be your single email delivery endpoint.

[Meakin Singh]

ok,but hw do i configure sendmail, i want to send mails(notifications) from HP UX server to windows clients

[Steven E. Protter]

Shalom,

vi /etc/rc.config.d/mailsvrs

Set the sendmail variable to 1 so the daemon auto starts.

/sbin/init.d/sendmail start

Configure sendmail.cf file to this:

DSrelayserver_hostname

replace relayserver_hostname with the hostname or IP address of a system that is configured to relay smtp traffic inside your organization.

SEP

[Meakin Singh]

sir,

i dont intend to send mails outside my organisation, plz tell me how do start mail from HP UX server to LAN windows based clients

[Meakin Singh]

Sir,

whenevr i try starting sendmail i get this error

/sbin/init.d/sendmail start

Mar 6 17:46:27 test sendmail[28950]: problem creating SMTP socket
Mar 6 17:47:12 test sendmail[28950]: NOQUEUE: SYSERR(root): opendaemonsocket: server SMTP socket wedged: exiting

[TTr]

> i dont intend to send mails outside my organisation, plz tell me how do start mail from HP UX server to LAN windows based clients

What SEP replied to you earlier is what is needed. Why did you assume he was telling you how to send email outside your organization?

If you have done what SEP suggested, you will need to talk to your exchange administrator to allow your EBS host to relay mail to the exchange system.

[Meakin Singh]

i get it sir

but what about the error i have mentioned

iam unable to start sendmail

[TTr]

> problem creating SMTP socket

Sendmail is probably running already. You need to stop it and restart it.

/sbin/init.d/sendmail stop

/sbin/init.d/sendmail start

[Mel Burslan]

As SEP previously said, in order to relay mail to MS Exchange servers, you have to be able to communicate with an exchange mail gateway, by means of configuring your smart relay host as

DSmail-relay-hostname

in your sendmail.cf file
Make sure you talk to the administrators of this mail relay host and have them add the ip address of your server to their authorized mail forwarders list (or whatever they are calling it). Otherwise, you will get mailbox not available messages every time you try to send a message

Hope this clears it.

[Meakin Singh]

hi, i cudnt get sendmail started
so i rebooted the server

now i am getting a diffrent error


# /sbin/init.d/sendmail start
/sbin/init.d/sendmail[54]: /etc/mail/sendmail.pid: Cannot find or open the file.
# Mar 6 23:07:43 test prngd[3051]: Will not touch /var/run/egd-pool: no socket

[Meakin Singh]

there is another problem
i have EBS running on this server

but ps -ef | grep applmgr returns no value

also

./a1 returns

ps -ef |grep java
0
ps -ef |grep oracle
0
ps -ef |grep LOCAL=NO
0
ps -ef |grep applmgr
0

but there are processes running on server

[TTr]

The sendmail might have started after you rebooted and no need to start it again.
"ps -ef |grep sendmail" will tell you what is running. Also check the bottom of /var/adm/syslog/mail.log for errors or status.

On another thought what is the SMTP requirement of EBS? Does it use the local SMTP or does it have an SMTP server configuration? If it is the latter, you could probably point EBS directly to the exchange SMTP relay server.

"ps -ef |more" look in detail fro running processes to see what is running.

[Meakin Singh]

sir,
sendmail is running but when i try to stop it, it gives the error

# /sbin/init.d/sendmail stop
/usr/sbin/killsm[7]: /etc/mail/sendmail.pid: cannot open

[Meakin Singh]

i guess realying is enabled by our mail admin , as emails from oracle(utl_smtp)package can be sent via this email server

the only problem is iam not able to send mail thru HP UX

i changed mailservs and sendmail.cf as suggested

and then use

sendmail user@domain.com

but i never recieve this email

[Meakin Singh]

to be more clear

my mail server IP is 192.168.0.221 , no idea abt its hostname

HP UX server IP , from here i want to send an email 192.168.0.93

my desktop ip , from where i want to recieve a mail sent from 192.168.0.93 is 192.168.0.33

[TTr]

> emails from oracle(utl_smtp)package can be sent via this email server

So as I suspected before the EBS is sending directly to the exchange server bypassing the local HP-UX sendmail.

Now let's verify the HP-UX sendmail.

Your desktop IP is not involved here. You still have to send email from HP-UX to the exchange server.
1) Did you put "DS192.168.0.221" in /etc/mail/sendmail.cf?

what is the output of
2) ps -ef | grep sendmail
3) ll /etc/mail/sendmail.cf
4) cat /etc/mail/sendmail.pid

[Meakin Singh]

1 yes, DS192.168.0.221 is in sendmail.cf

2 # ps -ef | grep sendmail
root 20721 15271 0 22:09:06 pts/tb 0:00 grep sendmail
root 9604 1 1 15:09:18 ? 0:17 sendmail: accepting connections on port 25

3 # ll /etc/mail/sendmail.cf
-r--r--r-- 1 bin bin 94367 Mar 6 17:56 /etc/mail/sendmail.cf

4 # cat /etc/mail/sendmail.pid
9604
/usr/sbin/sendmail -bd -q30m

[TTr]

Sendmail is running OK. Let's try sending a test email via your SMTP server.

echo TESTMAIL | sendmail -v email.address@yourdomain.com

where "email.address" is a valid recipient and "yourdomain.com" is your email domain in your 192.168.0.221 email system. Post the output of the above command

[Meakin Singh]

# echo TESTMAIL | sendmail -v meakin.singh@khannapaper.com
khannapaper.com: Name server timeout
meakin.singh@khannapaper.com... Transient parse error -- message queued for future delivery
meakin.singh@khannapaper.com... queued

[TTr]

You left out most of the output of the "...sendmail -v..." command and I can't tell where the problem is. You have to post the entire output. At least there is a name resolution issue possibly on your SMTP relay server and your HP-UX server sendmail.cf might need additional settings.

[Meakin Singh]

# echo TESTMAIL | sendmail -v meakin.singh@khannapaper.com
khannapaper.com: Name server timeout
meakin.singh@khannapaper.com... Transient parse error -- message queued for future delivery
meakin.singh@khannapaper.com... queued
#


sir,

this is the only output iam getting after running this command, is there some log generated ? id so , plaz tell me the location and i can post it here for your reference

[TTr]

Post the bottom few lines of /var/adm/syslog/mail.log
Also post the contents of /etc/resolv.conf. It looks like your server can not resolve any names (host/domain)
Do you have a DNS server in your environment?

[Meakin Singh]

Sir,

1. i dont have this file /etc/resolv.conf

2. i guess there is a DNS server , its ip is 192.168.0.120

3. lines from mail.log


Mar 9 17:07:01 test sendmail[3731]: RAA03729: to=root, ctladdr=root (0/3), delay=00:00:01, xdelay=00:00:01, mailer=local, stat=Sent
Mar 9 17:07:01 test sendmail[3736]: RAA03734: to=root, ctladdr=root (0/3), delay=00:00:00, xdelay=00:00:00, mailer=local, stat=Sent
Mar 9 17:07:01 test sendmail[3741]: RAA03739: to=root, ctladdr=root (0/3), delay=00:00:00, xdelay=00:00:00, mailer=local, stat=Sent
Mar 9 17:08:00 test sendmail[3834]: RAA03834: from=root, size=208, class=0, pri=30208, nrcpts=1, msgid=<200903091138.RAA03834@test.khannapaper.com>, relay=root@localhost
Mar 9 17:08:00 test sendmail[3836]: RAA03834: to=root, ctladdr=root (0/3), delay=00:00:00, xdelay=00:00:00, mailer=local, stat=Sent
Mar 9 17:08:54 test sendmail[3895]: RAA03895: to=meakin.singh@khannapaper.com, delay=00:00:00, mailer=relay, relay=192.168.0.221, stat=Transient parse error -- message queued for future delivery
Mar 9 17:08:54 test sendmail[3895]: RAA03895: from=root, size=9, class=0, pri=30009, nrcpts=1, msgid=<200903091138.RAA03895@test.khannapaper.com>, relay=root@localhost
Mar 9 17:08:54 test sendmail[3895]: RAA03895: to=meakin.singh@khannapaper.com, delay=00:00:00, mailer=relay, stat=queued
Mar 9 17:09:01 test sendmail[3917]: RAA03917: from=root, size=208, class=0, pri=30208, nrcpts=1, msgid=<200903091139.RAA03917@test.khannapaper.com>, relay=root@localhost
Mar 9 17:09:01 test sendmail[3919]: RAA03917: to=root, ctladdr=root (0/3), delay=00:00:00, xdelay=00:00:00, mailer=local, stat=Sent

[TTr · Accepted solution]

If you have a /etc/nsswitch.conf file post its contents.
If not, the try the following "cp /etc/nsswitch.files /etc/nsswitch.conf" If that fails, Create a /etc/nsswitch.conf" file and in it you put the following lines

passwd: files
group: files
hosts: files
ipnodes: files
services: files
networks: files
protocols: files
rpc: files
publickey: files
netgroup: files
automount: files
aliases: files

Then try the sendmail test again.
The /etc/nsswitch.conf file must be readable by everybody, permission mode 644.