I'll give a quick rundown of the problems we're having and see if anyone has any information that can help.

The server involved is a dedicated CPanel 9.72E-15/WHM 9.7.2 running RedHat Linux 7.3 i686.

The problems:

- The Exim SMTP server, while allowing connects to port 25, bounces any incoming message to a local valid address with both an "Unrouteable Domain" fatal error for the sending (remote) domain and with an "Unknown User" address for the local address. I checked for correct MX and A record resolution on the server with Dig, and the DNS is functioning correctly.

- The Exim SMTP server also reports invalid password for valid accounts on the server set to authenticate before sending mail. When authentication is turned off, Exim announces both "Verification failed for <localuser@localdomain.com>" and "Sender verify failed".

- The Redhat Package Manager (RPM) reports a "Segmentation Fault" when launched. It never even processes its own configuration before displaying the error. Redhat has no suggestion for that other than to re-install Red Hat from scratch (on a server? They must be joking).

If anyone has any suggestions or help to offer please do so quickly, as this server is populated by several people who can't use email at the moment. You may contact me by replying here, via email, or via AIM if the link actually works.

Thanks.

That "Segmentation Fault" does not sound very good. However it can be fixed but there is many causes for that. Also a possibliy perl is screwed up on the box which would cause email not to work correctly and rpm not to work correctly. There is also a chance your box is compromised If i was you id get the box wiped and a newer os installed but thats just me. Rh 7.3 is not supported and you probably got a whole slew of bugs, forexample the new exploit progeny patched today.

Originally posted by thelinuxguy
That "Segmentation Fault" does not sound very good. However it can be fixed but there is many causes for that. Also a possibliy perl is screwed up on the box which would cause email not to work correctly and rpm not to work correctly. There is also a chance your box is compromised If i was you id get the box wiped and a newer os installed but thats just me. Rh 7.3 is not supported and you probably got a whole slew of bugs, forexample the new exploit progeny patched today.

I'll look into the possibility of an issue with Perl. As for upgrading from RedHat 7.3, so far the company the server is rented from has resisted upgrading us to anything, so we'll see.

- J

In that case you may look at other companies. Moving cPanel is not hard if you find the right company to help if you are unsure.

There are many good companies out there.

I would recommend www.gnax.com and www.cpcservers.com

Originally posted by ddihosting
In that case you may look at other companies. Moving cPanel is not hard if you find the right company to help if you are unsure.

Thank you, that is a consideration for the future, but meanwhile I'd like to fix the email problem first. :)

Originally posted by mangacool
I'll give a quick rundown of the problems we're having and see if anyone has any information that can help.

The server involved is a dedicated CPanel 9.72E-15/WHM 9.7.2 running RedHat Linux 7.3 i686.

The problems:

- The Exim SMTP server, while allowing connects to port 25, bounces any incoming message to a local valid address with both an "Unrouteable Domain" fatal error for the sending (remote) domain and with an "Unknown User" address for the local address. I checked for correct MX and A record resolution on the server with Dig, and the DNS is functioning correctly.

- The Exim SMTP server also reports invalid password for valid accounts on the server set to authenticate before sending mail. When authentication is turned off, Exim announces both "Verification failed for <localuser@localdomain.com>" and "Sender verify failed".

- The Redhat Package Manager (RPM) reports a "Segmentation Fault" when launched. It never even processes its own configuration before displaying the error. Redhat has no suggestion for that other than to re-install Red Hat from scratch (on a server? They must be joking).

If anyone has any suggestions or help to offer please do so quickly, as this server is populated by several people who can't use email at the moment. You may contact me by replying here, via email, or via AIM if the link actually works.

Thanks.

A couple things to check on exim.

Make sure the domain is listed in the localdomains file

/etc/localdomains.

Next, check your tweak settings and see below is checked or unchecked..

Prevent the user 'nobody' from sending out mail to remote addresses (php and cgi scripts generally run as nobody if you are not using phpsuexec and suexec respectively.)

Next, You may want to downgrade from edge to a current or stable release. If that fails, run /scripts/exim4 from root. This usally clears up any issues with exim.

Originally posted by rmMark1
A couple things to check on exim.

Make sure the domain is listed in the localdomains file

/etc/localdomains.

Next, check your tweak settings and see below is checked or unchecked..

Prevent the user 'nobody' from sending out mail to remote addresses (php and cgi scripts generally run as nobody if you are not using phpsuexec and suexec respectively.)

Next, You may want to downgrade from edge to a current or stable release. If that fails, run /scripts/exim4 from root. This usally clears up any issues with exim.

All server domains are listed in /etc/localdomains. The second option will break all the PHP scripts on the server. And the /etc/exim4 script seems to be broken too - it exits complaining that some CPanel file doesn't exist:

(can't post the error, the board won't allow URL's)

Besides, with RPM broken it wouldn't matter if the file would download.

Is there an actual solution to this problem or should I ask people to begin backing up data and preparing for an extended server outage if it needs to be completely rebuilt?

To complicate matters, the people the server is rented from are refusing to service it any more due to the fact that we tried to repair it ourselves (we didn't really make it any worse, they just have some stupid policy) after they ignored the problem for 4 days.

Originally posted by mangacool
All server domains are listed in /etc/localdomains. The second option will break all the PHP scripts on the server. And the /etc/exim4 script seems to be broken too - it exits complaining that some CPanel file doesn't exist:

(can't post the error, the board won't allow URL's)

Besides, with RPM broken it wouldn't matter if the file would download.

Is there an actual solution to this problem or should I ask people to begin backing up data and preparing for an extended server outage if it needs to be completely rebuilt?

To complicate matters, the people the server is rented from are refusing to service it any more due to the fact that we tried to repair it ourselves (we didn't really make it any worse, they just have some stupid policy) after they ignored the problem for 4 days.

I would head over to cpanel.net and let them take a look.

http://cpanel.net/support/index.htm

Fill out the ticket with as much info. Then you may want to look for a new host that supports cpanel..

http://cpanel.net/dist.cgi

Originally posted by mangacool
All server domains are listed in /etc/localdomains. The second option will break all the PHP scripts on the server. And the /etc/exim4 script seems to be broken too - it exits complaining that some CPanel file doesn't exist:

(can't post the error, the board won't allow URL's)

Besides, with RPM broken it wouldn't matter if the file would download.

Is there an actual solution to this problem or should I ask people to begin backing up data and preparing for an extended server outage if it needs to be completely rebuilt?

To complicate matters, the people the server is rented from are refusing to service it any more due to the fact that we tried to repair it ourselves (we didn't really make it any worse, they just have some stupid policy) after they ignored the problem for 4 days.


There is always a possible fix. but theres not enough info given, the segfault could be a matter of things. You would need to look at an strace.

Originally posted by thelinuxguy
There is always a possible fix. but theres not enough info given, the segfault could be a matter of things. You would need to look at an strace.

Ok, aside from that: How about Exim? We can survive without RPM for a while but Exim is another matter. :)

Originally posted by mangacool
Ok, aside from that: How about Exim? We can survive without RPM for a while but Exim is another matter. :)


Looking at your cpanel version, exim might not have been installed correctly on update, therefor you need rpm back. And like i said before if perl is fubered then exim is not going to work.

Does your hostname resolve? Can you resolve thigns from the server?

Originally posted by thelinuxguy
Looking at your cpanel version, exim might not have been installed correctly on update, therefor you need rpm back. And like i said before if perl is fubered then exim is not going to work.

Does your hostname resolve? Can you resolve thigns from the server?

Yes, I tested it both from an outside PC and on the server using Dig. All DNS functions work, Apache works, PHP works, even CPanel and WHM seem to be working.

EDIT: And so far no one using space on the server has complained about Perl not working. Though two people have reported incoming mail being delivered at random in spite of "delivery failed" messages sent to the originator.

Originally posted by mangacool
Yes, I tested it both from an outside PC and on the server using Dig. All DNS functions work, Apache works, PHP works, even CPanel and WHM seem to be working.

EDIT: And so far no one using space on the server has complained about Perl not working. Though two people have reported incoming mail being delivered at random in spite of "delivery failed" messages sent to the originator.


Yeah but modules could be missing.

Originally posted by thelinuxguy
Yeah but modules could be missing.

That's another odd thing. About an hour ago Perl was recompiled, and gave no errors whatsoever. So if a module is missing it either is supposed to come with Exim or this particular CPanel version is defective and doesn't include what it should.

have you tried reinstalling the cpanel stuff with

/scripts/upcp --force?

Before he does that he needs to make sure RPM needs to be working.

Thinking about this it could be that the RPM problem caused some of the components to not be installed properly.

Originally posted by ddihosting
Before he does that he needs to make sure RPM needs to be working.

Thinking about this it could be that the RPM problem caused some of the components to not be installed properly.

Do you think it may help if someone out there who has a Redhat 7.3 i686 install can .tar.gz up RPM and send it to me?

Your bypassing the problem, chances are its not even rpm causing it, its something else on the server.

Originally posted by thelinuxguy
Your bypassing the problem, chances are its not even rpm causing it, its something else on the server.

Yes, we discovered that about 5 minutes ago. :) Any ideas? Anyone know what RPM is dependent on?

The problem has been fixed - RPM has been repaired and Exim successfully compiled - thanks to people who don't give up and a Swedish server admin who doesn't sleep. No thanks whatsoever to our rack host.

Originally posted by mangacool
The problem has been fixed - RPM has been repaired and Exim successfully compiled - thanks to people who don't give up and a Swedish server admin who doesn't sleep. No thanks whatsoever to our rack host.

You may want to look elsewhere if that is the support your getting. Glad you resolved your problem though...Good luck