Anyone else getting exim has failed, please contact the sysadmin ?

[Metro2]

I was just wondering if anyone else here has recently experienced the following error when trying to restart exim from WHM >> Main >> Restart Services >> Mail Server (Exim) :

---
Waiting for exim to restart....finished.

exim ([exim] ) running as mailnull with PID xxxxx (process table check method)

exim has failed, please contact the sysadmin.
---

I've just begun encountering this issue after running some manual commands suggested by cpanel in regard to "case ID # 55497". I'm not exactly sure how to find related documentation of "case ID # 55497" but that is what was referenced in a recent ticket I submitted to cpanel. A helpful cpanel tech suggested a couple of commands to run via ssh to patch a known bug, and since then I'm getting that error notice when I try to restart exim. Even running /scripts/eximup --force still results in "exim has failed, please contact the sysadmin.
Install Complete"

The strange thing is that it seems as though exim is working OK, but I simply cannot restart it without that failure notice / alert appearing.

I currently run WHM/cPanel 11.32.2 (build 18) "release" on RHEL 5 64-bit and up until the most recent update everything was going smooth.

I'd really like to know if anyone else has/is encountering this issue, so thank you for any replies.

Edit - PS - I do also have CSF/LFD/Mailscanner installed from configserver, but that has never created this issue in all the years I've been using CSF products with cPanel servers.

 

[Infopro]

I'm not exactly sure how to find related documentation of "case ID # 55497" but that is what was referenced in a recent ticket I submitted to cpanel.

That's an Internal ID number that you won't find anything on, it was put in your ticket for reference to your issue I would think. I've checked that Case myself and it doesn't read to me to be related to this at all. What was the ticket number you had/have open on this?

If that ticket is now closed, you should either reopen it or open a new one on this. That's the best way forward. Also, if you had not mentioned Mailscanner being installed, please do so on the new ticket. That may or may not be needed, but should be mentioned all the same I think.

I run Mailscanner as well on all my servers and if I restart EXIM from WHM I get the same message as you describe here. That said I do not think there's any issues with EXIM, or Mailscanner even. Both are online and working as expected.

It's always best to open a ticket for things like this as they can track any other tickets related and associate those having the same issue far more efficiently than a forum thread like this might.

HTH!

 

[Metro2]

Thank you very much for the feedback.

I do have a ticket open in regard to this issue, but the reason you're not associating the case ID is because of where this started.

I'll explain in decent detail here so that it makes sense as to why I started this thread, and you'll be able to easily put together the association with the case ID since you seem to be quite aware of it.

Here goes:

Around April 13th with the most recent "release" version of cPanel, my /usr/local/cpanel/logs/error_log starting showing tons of repeats of the following errors during cPanel Nightly Backups:

warn [pkgacct] Failed to open directory: /usr/local/cpanel/3rdparty/mailman/suspended.lists at /usr/local/cpanel/Cpanel/FileUtils/Match.pm line 26
Cpanel::FileUtils::Match::get_matching_files('/usr/local/cpanel/3rdparty/mailman/suspended.lists', '_(?:tip\\-username\\.com
Script::Pkgacct::script('Script::Pkgacct', 'username', '/backup/cpbackup/daily', 'backup') called at /usr/local/cpanel/scripts/pkgacct line 90

and

warn [featureadmin] User file '/var/cpanel/users/root' is empty or non-existent. at /usr/local/cpanel/Cpanel/Config/LoadCpUserFile.pm line 115
Cpanel::Config::LoadCpUserFile::_load('root', undef, 0) called at /usr/local/cpanel/Cpanel/Config/LoadCpUserFile.pm line 24
Cpanel::Config::LoadCpUserFile::load('root') called at bin/featureadmin line 38

Either of those error paragraphs will repeat in /usr/local/cpanel/logs/error_log during nightly backup for as many customer accounts as there are on each server.

In other words - if I have 200 accounts on a server, then every night /usr/local/cpanel/logs/error_log will show 200 instances of:

warn [pkgacct] Failed to open directory: /usr/local/cpanel/3rdparty/mailman/suspended.lists at /usr/local/cpanel/Cpanel/FileUtils/Match.pm line 26
Cpanel::FileUtils::Match::get_matching_files('/usr/local/cpanel/3rdparty/mailman/suspended.lists', '_(?:tip\\-username\\.com
Script::Pkgacct::script('Script::Pkgacct', 'username', '/backup/cpbackup/daily', 'backup') called at /usr/local/cpanel/scripts/pkgacct line 90

Note - this issue didn't start occurring on my servers until Release 11.32.2 (build 18)

So I submitted a ticket about that to cpanel support 2 days ago on April 19th.

Today I received a response from a cpanel tech who investigated and he said:

This issue has recently been reported to our developers, and
has been fixed in case ID # 55497. The case has recently been
tagged with "11.32.3.1", indicating that the fix will be released in
that version of cPanel. There is no ETA on when that version will
be released at this time.

In the meantime, you can resolve this behavior using these commands:

[root@cpj ~]# mkdir /usr/local/cpanel/3rdparty/mailman/suspended.lists
[root@cpj ~]# chown mailman.mailman /usr/local/cpanel/3rdparty/mailman/suspended.lists
[root@cpj ~]# chmod 2775 /usr/local/cpanel/3rdparty/mailman/suspended.lists

You could probably get away with just running the first command, but I've
included the others so that the directory will have the correct user:group
ownership and permissions, which could help prevent issues if you ever
decide to enable Mailman.

I ran the suggested commands provided by the cpanel technician as directed.

Now if I attempt to restart exim from WHM or SSH, I receive the error

"exim has failed, please contact the sysadmin"

Like your experience (Infopro), my exim and mailscanner both appear to still be operating OK.

BUT, before the most recent cPanel update:

A) The errors in /usr/local/cpanel/logs/error_log (that cpanel has documented as a true bug slated to be fixed in a future release) never occured / generated before.

and

B) I never had a problem restarting exim from WHM

So that's the whole of it - a bug (known by cpanel devs) in the latest release of cpanel is generating errors (and a fix is coming in future release) but in the meantime I ran the manual fix commands as instructed, and then became concerned about the fact that suddenly I can no longer restart exim via WHM or SSH, and so I posted this thread.

While you're definitely correct that submitting tickets directly to cpanel about such issues (which I did 2 days ago) sometimes, as you may well know, it helps to discuss trickle-down issues with the community here as well. Especially when you want to know if you're only experiencing a certain part of an error after running certain suggested commands, or if many people are experiencing the problem in general.

Although we both know that Mailscanner is not the culprit, as cpanel has already acknowledged the bug in the case ID that you are familiar with and has nothing to do with mailscanner, I will surely take your advice and add to the ticket the fact that I do have it installed.

Thanks very much for the response!

 

[Infopro]

In the meantime, you can resolve this behavior using these commands:

If I read that right, he instructed you to create a directory, give it an owner and set permissions on it. So that bug is, I guess, a missing directory that generates an error. That'll be fixed I'm sure.

Restarting EXIM and getting the message we both get, I would think, is unrelated.

I would only add here that over the years I've seen these sort of false positive messages if you will, before. Go to WHM and restart SSH, right now. Mine gives me that same message about the service failing. I'm going to guess that yours will as well.

Is this a defect? Yes I think so. Is it problematic to me and giving me grief? No. As long as that service starts back up, the message that the system spits to my browser, while it should give a proper response and may not, means less to me.

I don't want to sound like I'm excusing this message itself, I'm not. But the way that message is spit to my browser IMHO, might need some tweaking, yes.


If creating that directory solved those backup errors, great. If you've reported that message from EXIM in a ticket, great. Be sure to provide login details so cPanel Support can take a closer look. I do not think they're related though. Just my opinion of course, ask in your ticket and get a proper response from someone who can.

 

[Metro2]

I wasn't looking for anyone to excuse the issue. I'm here to discuss it and find out if anyone else has begun experiencing it. (That is one of the things this forum is for, right? )

And so far it's working out good - you've confirmed for me that you're experiencing the same bug. Very helpful to know that I'm not the only one.

Also worth noting - In the past 10 years of using WHM/cPanel on many servers over those years, I have never experienced any of the errors discussed in this thread until Release 11.32.2 (build 18).

Are the errors killing me? No.

Are they significant and worth investigating / addressing? Absolutely! They're newly discovered "bugs".

I hope that this is the right place to discuss them.

I'm getting a vibe from you that you and I just don't click. It's happened before, but you probably don't remember me because you do a lot more posting / corresponding on here than I do.

If you don't like me or my tone, that's fine. But don't sh*tbag every thread / post I make.

 

[Infopro]

I remember every single person I've ever dealt with on these forums, including you. And, I have zero issues with you at all.

If you don't like me or my tone, that's fine. But don't sh*tbag every thread / post I make.

You seem to think there's an issue here where this is none. I'm hoping to assist you the best I can. If you'd rather I didn't, please let me know and I'll add you to my blocked list and will no longer see any of your posts anywhere on these forums.

I don't mind helping you out when I can. My apologies if you think that's shitbagging you. Whatever that means..

 

[Metro2]

I suppose it's just me then. I feel passive-agressive condescension mixed in with your replies to me, things like be-littling the fact that I brought up an actual bug here on the forum and you making it seem like I shouldn't have bothered because it's not causing you any problems.

Bouncing back and forth between two threads with you right now, your replies have made me feel uncomfortable and I feel that you belittle my statements and opinions.

I'll try to use more specific terms like the above from here forth so that you don't have to end your replies to me with things like "Whatever that means..."

 

[cPanelTristan]

Are they significant and worth investigating / addressing? Absolutely! They're newly discovered "bugs".

I hope that this is the right place to discuss them.

I want to note that bug reports when a product defect is suspected should be posted to http://go.cpanel.net/bugs (or using the link at the top of the forum) in all instances. Bug reports submitted via the forums are redirected per our request to that area of our support desk.

Now, even if mailscanner and exim are running properly, it's quite possible that some change in the detection of exim running changed in 11.32 recently and that mailscanner is conflicting with that process. I would suggest temporarily uninstalling mailscanner just to test if that's the case. If it then starts properly working, you can conclusively know it is mailscanner that is the issue for the service restart check. I don't run mailscanner on my system and I'm not experiencing this issue. You both are running mailscanner and are experiencing it. That isn't conclusive evidence of the cause unless mailscanner is uninstalled and it works fine at that point.

 

[Metro2]

Thanks for the reply and advice Tristan.

Here's the funny thing - without making any changes or removing Mailscanner, suddenly without explanation the problem stopped on all but one of my servers. I can restart exim from WHM on all but one server now (and they're basically identical configurations, modern servers provisioned this year).

Maybe this is a coincidence, but the exim restart problem went away the day after I posted in my ticket about it and a cPanel level III tech responded with:

We have case 58442 open on service restarts showing failure even though they succeeded. The problem there is that the service check is happening too quickly after the restart is issued, not giving the restart time enough to happen. It as well has been fixed in the upcoming 11.32.3 branch.

So we've ruled-out Mailscanner I think (between the fact that A) the problem went away on it's own within 48 hours of starting a ticket and B) I never disabled or removed or changed Mailscanner and C) an experienced cPanel tech has explained that the issue stems from "the service check is happening too quickly after the restart".

There's still some mystery shrouding this (fixed itself on 2 out of 3 of my servers, some people encounter the issue, some people don't), but at least now by bringing it up in a support ticket we get an answer, a reason for it happening, and a case number letting us know a fix is ccoming someday.

 

[Metro2]

On a side-note - my *sincere* apologies to the community for almost de-railing this thread over an issue I thought I had with Infopro. I'm not going to get into detail, but for whatever reason (perhaps me being too sensitive in some case) his replies to some posts I was making in a separate thread at the same time as this one were really rubbing me the wrong way and I guess I let that emotion spill over into this thread. I still think that Infopro and I just don't "mesh" sometimes and that's just something that happens on public forums, but:

Please note that regardless of any friction I've always respected Infopro and I think he is a very helpful part of the community. I wish I could explain why on certain occasions I've taken something a bit too personally from him, but this thread really wasn't the problem and I apologize to everyone (including Infopro) for kind of losing my cool a little.

An no, I was not prompted by anyone else to post this apology to the community. This is on my own accord because I think it's important to ackowledge it when I've made an un-productive post or diluted a thread, and attone for it.

 

[Infopro]

That's very kind of you to take the time to post that comment.

I wish I could explain why on certain occasions I've taken something a bit too personally from him..

You wouldn't be the first, nor the last here on this, trust me on that. I have horrible typing skills that do read different at times than what I hoped for when I look back. I spend a lot of time on many forums and sometimes pressed for time, I'll just get right to the thing instead of concerning myself with slowing down enough to be a bit more Netiquette friendly, as I should.

I'm actually a really friendly person, it's this text on the page that makes me sound otherwise. I'm not sure a mean person could do what I do every day, or would even want to.

My apologies for any misunderstandings here, or over there in the other thread.