F20 Where's my system mail?

[Arthur Dent]

Hello all,

I have read this page:
https://fedoraproject.org/wiki/Changes/NoDefaultSendmail

Having upgraded to F20 the new lack of an MTA caused me no end of
trouble with my fetchmail->procmail->dovecot setup, but I think I have
got that sorted now. I had a brief flirtation with postfix, but gave
that up as a bad job when it started bouncing all my mail.

In truth I am quite happy not to have an MTA and risk accidentally
spamming the interweb, but I now find I am lacking one thing...

My system mail (and output from my user cron jobs) used to find it's way
into procmail from which I could direct it into mailboxes to be read at
my leisure.

Without a functioning MTA how can I now achieve the same result?
Although I can ssh into the box, and I do often check logs etc, it is
reassuring to know when something has happened I would get an email, and
the regular output in email-form from logwatch was something I would
check every day.

Can I get this functionality back without the pain of configuring
postfix or (shudder) sendmail? If so how?

Thanks in advance.

Mark

[Arthur Dent]

On Tue, 2014-04-08 at 22:22 +0100, Arthur Dent wrote:
> Hello all,
>
> I have read this page:
> https://fedoraproject.org/wiki/Changes/NoDefaultSendmail
>
> Having upgraded to F20 the new lack of an MTA caused me no end of
> trouble with my fetchmail->procmail->dovecot setup, but I think I have
> got that sorted now. I had a brief flirtation with postfix, but gave
> that up as a bad job when it started bouncing all my mail.
>
> In truth I am quite happy not to have an MTA and risk accidentally
> spamming the interweb, but I now find I am lacking one thing...
>
> My system mail (and output from my user cron jobs) used to find it's way
> into procmail from which I could direct it into mailboxes to be read at
> my leisure.
>
> Without a functioning MTA how can I now achieve the same result?
> Although I can ssh into the box, and I do often check logs etc, it is
> reassuring to know when something has happened I would get an email, and
> the regular output in email-form from logwatch was something I would
> check every day.
>
> Can I get this functionality back without the pain of configuring
> postfix or (shudder) sendmail? If so how?
>
> Thanks in advance.

Gentle bump...

Does everyone here now do without email alerts, or does everyone just
install postfix or some such?

I would really like to know if there is a workaround for me...

Thanks again

Mark

[Joonas Sarajärvi]

2014-04-10 17:08 GMT+03:00 Arthur Dent :
> Does everyone here now do without email alerts, or does everyone just
> install postfix or some such?
>
> I would really like to know if there is a workaround for me...
>
> Thanks again
>
> Mark

My impression is that an MTA is currently not installed by default
because it was thought that most users do not use the things it
provides. However, since you seem to actually use functionality that
requires an MTA to be present, I think you should just install some
MTA of your choice and keep using the system mail functionality.

-Joonas

[Arthur Dent]

On Thu, 2014-04-10 at 17:13 +0300, Joonas Sarajärvi wrote:
> 2014-04-10 17:08 GMT+03:00 Arthur Dent :
> >
>
> My impression is that an MTA is currently not installed by default
> because it was thought that most users do not use the things it
> provides. However, since you seem to actually use functionality that
> requires an MTA to be present, I think you should just install some
> MTA of your choice and keep using the system mail functionality.

Thanks for helping.

My problems are twofold:
1) When I tried to install postfix I found it was bouncing ALL my mail!
(I use fetchmail->procmail->dovecot on this small family server - just
receiving / filtering - not sending).
2) now that Fedora have removed the MTA as default, does it even
generate system messages and cron output messages anymore?

Mark

[Steve Searle]

Around 03:24pm on Thursday, April 10, 2014 (UK time), Arthur Dent wrote:
> now that Fedora have removed the MTA as default, does it even
>
> generate system messages and cron output messages anymore?

Yes it does. I install and configure sendmail on both my Fedora
workstations and do receive all the expected system messages as I did
before the default installation of sendmail was removed from Fedora.

I my opintion the removal of sendmail was a wrong decision, but this has
been debated to death on this list before.

Steve

[Arthur Dent]

On Thu, 2014-04-10 at 15:51 +0100, Steve Searle wrote:
> Around 03:24pm on Thursday, April 10, 2014 (UK time), Arthur Dent wrote:
> > now that Fedora have removed the MTA as default, does it even
> >
> > generate system messages and cron output messages anymore?
>
> Yes it does. I install and configure sendmail on both my Fedora
> workstations and do receive all the expected system messages as I did
> before the default installation of sendmail was removed from Fedora.
>
> I my opintion the removal of sendmail was a wrong decision, but this has
> been debated to death on this list before.

OK - So if just do "yum install sendmail" will everything work as
before? Or will I have to spend a fortune on coffee and pizzas as I sit
up night after night trying to configure it?

[Steve Searle]

Around 04:47pm on Thursday, April 10, 2014 (UK time), Arthur Dent wrote:
> OK - So if just do "yum install sendmail" will everything work as
> before? Or will I have to spend a fortune on coffee and pizzas as I sit
> up night after night trying to configure it?

Not a fortune, but you need to tinker a little. I have my own CentOS
server on my network which is my actual mail server, and my instructions
for each time I install Fedora are as follows:

Edit /etc/mail/sendmail.mc and add the following lines, typically
  where there commented out template is. The MAIL_HUB line follows the
  'SMART_HOST line and causes mail that would have otherwise have been
  sent to a local user to be sent to the network's mail server. The
  EXPOSED_USER line needs to be deleted or commented out to allow root
  to send masqueraded emails.

...
    define(`SMART_HOST', `flounder.stevesearle.com')
    define(`MAIL_HUB', `relay:flounder.stevesearle.com')
    ...
    dnl #EXPOSED_USER(`root')dnl
    ...
    MASQUERADE_AS(`stevesearle.com')
    ...
    FEATURE(masquerade_envelope)
    ...

Regenerate /etc/mail/sendmail.cf. The sendmail-cf package must be
  installed for this, and no warnings are displayed if it isn't installed.

# make -C /etc/mail

Restart sendmail.

# systemctl restart sendmail.service

Edit /etc/aliases so that root's email is redirected to my account.

...
    root:    steve

Run newaliases for the change to be shown through sendmail.

# newaliases

I assume you need to have your SMART_HOST and MAIL_HUB lines refering to
your ISP's mail server if you want the emails passed on to there.

Steve

[Arthur Dent]

On Thu, 2014-04-10 at 17:42 +0100, Steve Searle wrote:
> Around 04:47pm on Thursday, April 10, 2014 (UK time), Arthur Dent wrote:
> > OK - So if just do "yum install sendmail" will everything work as
> > before? Or will I have to spend a fortune on coffee and pizzas as I sit
> > up night after night trying to configure it?
>
> Not a fortune, but you need to tinker a little. I have my own CentOS
> server on my network which is my actual mail server, and my instructions
> for each time I install Fedora are as follows:

Apologies for the slow response, but I had to attend a funeral out of
town and have been away from my computer for the last few days.

To recap:
I use fetchmail->procmail->dovecot to process incoming mails. I use the
smtp client built into Evolution to send mails (with one exception - I
use squirrelmail to access my dovecot server remotely and, in installing
squirrelmail via yum I noticed it pulled in exim too - so I now have
exim on my computer it seems).

All I really want is for internal system mails from logwatch and cron
jobs to be poked into procmail INTERNALLY. I see no need for them to go
outside into the big wide world. I am very nervous of attempting to
configure sendmail / postfix / exim because I am not really well versed
in MTA configuration, and when I tried to install postfix I think it
sent bounce messages to every incoming mail. I am terrified of becoming
listed as a spam source (I think my reputation may already be damaged
from an earlier misconfiguration).

Seeing that I had exim installed I looked into some setup guides. I
found this:
http://newbiedoc.sourceforge.net/networking/exim.html
which seems to be exactly what I want - internal only mail.
The problem is that it relies on a too called eximconfig which does not
appear to be available with the Fedora package.

Can anybody help me achieve what I want - either avoid the need for an
MTA altogether, or configure a simple MTA to process internal mail
internally?

Thanks

Mark

[Suvayu Ali]

On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
> Can anybody help me achieve what I want - either avoid the need for an
> MTA altogether, or configure a simple MTA to process internal mail
> internally?

If you are not sending emails over the internet, no configuration is
required.  Just install an MTA like sendmail/postfix/exim and turn on
the service with default configurations.  That should take care of all
intenal mail.

I read system mail with mutt; my only setup is the following line in
/etc/aliases:

root:		<myusername>

Hope this helps,

[Arthur Dent]

On Sun, 2014-04-13 at 12:43 +0200, Suvayu Ali wrote:
> On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
> > Can anybody help me achieve what I want - either avoid the need for an
> > MTA altogether, or configure a simple MTA to process internal mail
> > internally?
>
> If you are not sending emails over the internet, no configuration is
> required.  Just install an MTA like sendmail/postfix/exim and turn on
> the service with default configurations.  That should take care of all
> intenal mail.
>
> I read system mail with mutt; my only setup is the following line in
> /etc/aliases:
>
> root:		<myusername>
>
> Hope this helps,

That is fantastically helpful - thank you!

This means that I get get system mail for root (my main concern) but it
doesn't seen to give me cron output for jobs run as my normal user.

Can I achieve something similar for that? If so how?

Thanks

Mark

[Suvayu Ali]

On Sun, Apr 13, 2014 at 02:28:38PM +0100, Arthur Dent wrote:
> On Sun, 2014-04-13 at 12:43 +0200, Suvayu Ali wrote:
> > On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
> > > Can anybody help me achieve what I want - either avoid the need for an
> > > MTA altogether, or configure a simple MTA to process internal mail
> > > internally?
> >
> > If you are not sending emails over the internet, no configuration is
> > required.  Just install an MTA like sendmail/postfix/exim and turn on
> > the service with default configurations.  That should take care of all
> > intenal mail.
> >
> > I read system mail with mutt; my only setup is the following line in
> > /etc/aliases:
> >
> > root:		<myusername>
> >
> > Hope this helps,
>
> That is fantastically helpful - thank you!
>
> This means that I get get system mail for root (my main concern) but it
> doesn't seen to give me cron output for jobs run as my normal user.
>
> Can I achieve something similar for that? If so how?

That requires nothing at all.  All system mail will be delivered to
/var/spool/mail/<username> without any configuration.

[Arthur Dent]

On Sun, 2014-04-13 at 15:43 +0200, Suvayu Ali wrote:
> On Sun, Apr 13, 2014 at 02:28:38PM +0100, Arthur Dent wrote:
> > On Sun, 2014-04-13 at 12:43 +0200, Suvayu Ali wrote:
> > > On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
> > This means that I get get system mail for root (my main concern) but it
> > doesn't seen to give me cron output for jobs run as my normal user.
> >
> > Can I achieve something similar for that? If so how?
>
> That requires nothing at all.  All system mail will be delivered to
> /var/spool/mail/<username> without any configuration.

Oh Dear. That means I have a problem somewhere.
I have created a simple script:

#!/bin/bash
#-----------------------------------------------------------------
# Script to check for the existence of the file ~/testfile
# If the file is present cron should send an email to warn of that fact.
#
if [ -f /home/mark/testfile ]
then
    echo Warning! The file Tesfile exists!
fi

Running this from the command line produces the expected output. Running
it from cron (as my user - mark) does nothing.

What have I done wrong?

Thanks again.

Mark

[Suvayu Ali]

On Sun, Apr 13, 2014 at 03:38:57PM +0100, Arthur Dent wrote:
> On Sun, 2014-04-13 at 15:43 +0200, Suvayu Ali wrote:
> > On Sun, Apr 13, 2014 at 02:28:38PM +0100, Arthur Dent wrote:
> > > On Sun, 2014-04-13 at 12:43 +0200, Suvayu Ali wrote:
> > > > On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
> > >
> > > This means that I get get system mail for root (my main concern) but it
> > > doesn't seen to give me cron output for jobs run as my normal user.
> > >
> > > Can I achieve something similar for that? If so how?
> >
> > That requires nothing at all.  All system mail will be delivered to
> > /var/spool/mail/<username> without any configuration.
>
> Oh Dear. That means I have a problem somewhere.
> I have created a simple script:
>
> #!/bin/bash
> #-----------------------------------------------------------------
> # Script to check for the existence of the file ~/testfile
> # If the file is present cron should send an email to warn of that fact.
> #
> if [ -f /home/mark/testfile ]
> then
>     echo Warning! The file Tesfile exists!
> fi
>
> Running this from the command line produces the expected output. Running
> it from cron (as my user - mark) does nothing.
>
> What have I done wrong?

I'm not sure, I tried the following cron entry to test:

[[ -f $HOME/dummy ]] || echo does not exist

I received mail like this:

Subject: Cron  [[ -f $HOME/dummy ]] || echo does not exist

does not exist

Are you sure your MTA is running properly?  What does `systemctl status'
tell you?

Hope this helps,

[Arthur Dent]

On Sun, 2014-04-13 at 17:12 +0200, Suvayu Ali wrote:
> On Sun, Apr 13, 2014 at 03:38:57PM +0100, Arthur Dent wrote:
> > On Sun, 2014-04-13 at 15:43 +0200, Suvayu Ali wrote:
> > > On Sun, Apr 13, 2014 at 02:28:38PM +0100, Arthur Dent wrote:
> > > > On Sun, 2014-04-13 at 12:43 +0200, Suvayu Ali wrote:
> > > > > On Sun, Apr 13, 2014 at 10:54:18AM +0100, Arthur Dent wrote:
>
> Are you sure your MTA is running properly?  What does `systemctl status'
> tell you?

Well I'm pretty sure my MTA is running. I get system emails for root (I
have just restarted fail2ban.service and that sends an email to confirm
it has started.

Here is the output of systemctl status:

# systemctl status exim.service 
exim.service - Exim Mail Transport Agent
   Loaded: loaded (/usr/lib/systemd/system/exim.service; enabled)
   Active: active (running) since Sun 2014-04-13 12:08:25 BST; 4h 18min ago
 Main PID: 15449 (exim)
   CGroup: /system.slice/exim.service
           └─15449 /usr/sbin/exim -bd -q1h

Apr 13 12:08:20 Zaphod systemd[1]: Starting Exim Mail Transport Agent...
Apr 13 12:08:21 Zaphod exim-gen-cert[15440]: Generating exim certificate: /usr/libexec/exim-gen-cert: line 25: success:... found
Apr 13 12:08:25 Zaphod systemd[1]: Started Exim Mail Transport Agent.
Hint: Some lines were ellipsized, use -l to show in full.

It's just my own cron jobs that don't come through.

I know that the job has worked:
# cat /var/log/cron
...
Apr 13 15:35:07 Zaphod crontab[17561]: (mark) BEGIN EDIT (mark)
Apr 13 15:35:39 Zaphod crontab[17561]: (mark) REPLACE (mark)
Apr 13 15:35:39 Zaphod crontab[17561]: (mark) END EDIT (mark)
Apr 13 15:36:01 Zaphod crond[573]: (mark) RELOAD (/var/spool/cron/mark)
Apr 13 15:36:01 Zaphod CROND[17568]: (mark) CMD (~/scripts/croncheck.sh)
Apr 13 15:36:01 Zaphod CROND[17564]: (mark) CMDOUT (Warning! The file Tesfile exists!)

[Arthur Dent]

Small update:
Now that root mail is working, I have this morning received in my email
the output of logwatch - which is great - I notice that (unlike with my
previous Fedora installs) it includes a "cron" section. In this I can
see the output of the cron jobs run as my user (mark) as well as those
run by root.

--------------------- Cron Begin ------------------------ 
(snip...)
 CMDOUT (Warning! The file Tesfile exists!)
 CMDOUT (Warning! The file Tesfile exists!)
 CMDOUT (Warning! The file Tesfile exists!)
 CMDOUT (Warning! The file Tesfile exists!)
 CMDOUT (Warning! The file Tesfile exists!)
 CMDOUT (Warning! The file Tesfile exists!) 
(snip...)
 ---------------------- Cron End -------------------------

This is better than nothing, but I would still prefer to get the output
of cron jobs put in my user mail file.

The test example I gave yesterday is in fact based on a real-life case.
I have a job that runs hourly from cron. Occasionally this process fails
(for reasons I won't bore the list with) and leaves a lock-file in place
- a will then refuse to run again until someone (me) has cleaned up
after it. The only way I even know that this has happened is accuse I
have another little hourly cron job that checks for the presence of the
lock-file and reports only if it exists.

If I have to rely on the daily logwatch output it could conceivably be
up to 23 hours before I even know that the job is not working...

So - how to get my user system-mail back?

Thanks for all the help so far...

Mark

[Fred Smith]

On Mon, Apr 14, 2014 at 09:20:09AM +0100, Arthur Dent wrote:
<snip>
> The test example I gave yesterday is in fact based on a real-life case.
> I have a job that runs hourly from cron. Occasionally this process fails
> (for reasons I won't bore the list with) and leaves a lock-file in place
> a will then refuse to run again until someone (me) has cleaned up
>
> after it. The only way I even know that this has happened is accuse I
> have another little hourly cron job that checks for the presence of the
> lock-file and reports only if it exists.
>
> If I have to rely on the daily logwatch output it could conceivably be
> up to 23 hours before I even know that the job is not working...

I've used, in the past, a mechanism where the program controls its own
lock file, so it has control over what goes inside the file. (I dunno if
your app is simlar, or if it uses some system services to manage lockfiles.)

When starting up:
1. see if the lockfile exists
1a. Exists: open it, read its contents, which should be the PID of the
    program that created it. see if a process of that PID is still running. 
    If so, silently exit. if the pid doesn't exist, delete the lock file
    and go to step 2.
1b. if no lockfile, start normally.
2. normal start: create the lockfile, write your own PID into it.
3. program runs.
4. when program finishes, it removes the lockfile.

This lets the program determine, whenever it starts up, if it's a bogus lock
file, or if it isn't. If it no longer refers to a running process, it can
simply be deleted without causing any undue delays.

[Arthur Dent]

On Mon, 2014-04-14 at 09:10 -0400, Fred Smith wrote:
> On Mon, Apr 14, 2014 at 09:20:09AM +0100, Arthur Dent wrote:
>
> <snip>
> > The test example I gave yesterday is in fact based on a real-life case.
> > I have a job that runs hourly from cron. Occasionally this process fails
> > (for reasons I won't bore the list with) and leaves a lock-file in place
> > - a will then refuse to run again until someone (me) has cleaned up
> > after it. The only way I even know that this has happened is accuse I
> > have another little hourly cron job that checks for the presence of the
> > lock-file and reports only if it exists.
> > 
> > If I have to rely on the daily logwatch output it could conceivably be
> > up to 23 hours before I even know that the job is not working...
>
> I've used, in the past, a mechanism where the program controls its own
> lock file, so it has control over what goes inside the file. (I dunno if
> your app is simlar, or if it uses some system services to manage lockfiles.)
>
> When starting up:
> see if the lockfile exists
>
> 1a. Exists: open it, read its contents, which should be the PID of the
>     program that created it. see if a process of that PID is still running. 
>     If so, silently exit. if the pid doesn't exist, delete the lock file
>     and go to step 2.
> 1b. if no lockfile, start normally.
> 2. normal start: create the lockfile, write your own PID into it.
> 3. program runs.
> 4. when program finishes, it removes the lockfile.
>
> This lets the program determine, whenever it starts up, if it's a bogus lock
> file, or if it isn't. If it no longer refers to a running process, it can
> simply be deleted without causing any undue delays.

Thanks for that - I am going to look into fixing that at some point in
the future, but at the moment my workaround will suffice. It's not just
for that problem that I want to see cron output in email form. I have
lots of jobs run as my user (mark) and I want the email to to know when
they have completed etc.

I am tearing my hair out over this. I have read previous threads on the
subject, but most seem to relate to getting root mail. That works now
(thanks to Suvayu Ali) - putting:
# Person who should get root's mail
root:		mark
into /etc/aliases did the trick for root cron jobs.
But I still don't get cron output for user mark.

I have MAILTO=mark at the top of my crontab file.

According to /var/log/cron the job has run and has produced output:
Apr 14 13:58:02 Zaphod CROND[8001]: (mark) CMD (~/scripts/croncheck.sh)
Apr 14 13:58:02 Zaphod CROND[7999]: (mark) CMDOUT (Please Note: That file does not exist)

Nothing appears in /var/spool/mail/mark though.

The exim logs show no evidence of having been bothered by cron...
2014-04-14 12:44:05 1WZfIy-0001xD-Mb <= [e-mail] U=root P=local S=1988
2014-04-14 12:44:06 1WZfIy-0001xD-Mb => mark  R=procmail T=procmail
2014-04-14 12:44:06 1WZfIy-0001xD-Mb Completed
2014-04-14 13:08:35 Start queue run: pid=7745
2014-04-14 13:08:35 End queue run: pid=7745
2014-04-14 14:08:35 Start queue run: pid=8149
2014-04-14 14:08:35 End queue run: pid=8149

Note that the entry at 12:44 was the output from fail2ban which is
mailed to root (and thence to Mark). I then get that email.

So where does the output from cron jobs for normal users go?

[Arthur Dent]

> I am tearing my hair out over this. I have read previous threads on the
> subject, but most seem to relate to getting root mail. That works now
> (thanks to Suvayu Ali) - putting:
> # Person who should get root's mail
> root:		mark
> into /etc/aliases did the trick for root cron jobs.
> But I still don't get cron output for user mark.
>
> I have MAILTO=mark at the top of my crontab file.
>
> According to /var/log/cron the job has run and has produced output:
> Apr 14 13:58:02 Zaphod CROND[8001]: (mark) CMD (~/scripts/croncheck.sh)
> Apr 14 13:58:02 Zaphod CROND[7999]: (mark) CMDOUT (Please Note: That file does not exist)
>
> Nothing appears in /var/spool/mail/mark though.
>
> The exim logs show no evidence of having been bothered by cron...
> 2014-04-14 12:44:05 1WZfIy-0001xD-Mb <= [e-mail] U=root P=local S=1988
> 2014-04-14 12:44:06 1WZfIy-0001xD-Mb => mark  R=procmail T=procmail
> 2014-04-14 12:44:06 1WZfIy-0001xD-Mb Completed
> 2014-04-14 13:08:35 Start queue run: pid=7745
> 2014-04-14 13:08:35 End queue run: pid=7745
> 2014-04-14 14:08:35 Start queue run: pid=8149
> 2014-04-14 14:08:35 End queue run: pid=8149
>
> Note that the entry at 12:44 was the output from fail2ban which is
> mailed to root (and thence to Mark). I then get that email.
>
> So where does the output from cron jobs for normal users go?

Another update:
I do have msmtp installed on this system in order to be able to mail
from php scripts. I have temporarily disabled it by moving .msmtprc out
of the way.

When I log in as another user on the system (Let's call her Mary) I can
do:
$ mail mark
Subject: Test Message
Just testing...
[CTRL-d]
EOT

and the message appears in /var/spool/mail/mark

But when I log in as Mark I get this:
$ mail mary
Subject: Testing internal messaging
Testing internal email
[CTRL-d]
EOT
[mark@Zaphod]$ send-mail: account default not found: no configuration file available

Is this significant?

[Suvayu Ali]

Hi Arthur,

Sorry about my late response.

On Mon, Apr 14, 2014 at 04:51:38PM +0100, Arthur Dent wrote:
> Another update:
> I do have msmtp installed on this system in order to be able to mail
> from php scripts. I have temporarily disabled it by moving .msmtprc out
> of the way.
>
> When I log in as another user on the system (Let's call her Mary) I can
> do:
> $ mail mark
> Subject: Test Message
> Just testing...
> [CTRL-d]
> EOT
>
> and the message appears in /var/spool/mail/mark
>
> But when I log in as Mark I get this:
> $ mail mary
> Subject: Testing internal messaging
> Testing internal email
> [CTRL-d]
> EOT
> [mark@Zaphod]$ send-mail: account default not found: no configuration file available

It looks like this is the crux of your problem.  I presume you checked
mary gets mails for her cron jobs?

That said, I do not know enough about system mail to point you to
possible sources of your problem, however I would start by running mail
with the verbose flag.
> Is this significant?

I would say, definitely so.

Hope this helps,

[Arthur Dent]

On Mon, 2014-04-14 at 20:25 +0200, Suvayu Ali wrote:
> > [mark@Zaphod]$ send-mail: account default not found: no configuration file available
>
> It looks like this is the crux of your problem.  I presume you checked
> mary gets mails for her cron jobs?

Well I think we're narrowing down the problem here. The answer to that
question is NO!
And interestingly NEITHER DOES ROOT!

It seems that fixing /etc/aliases now allows me to get SYSTEM messages
(e.g. logwatch, fail2ban stops/starts etc.) but NOT cron output.

So the problem seems to be that cron is somehow not configured to send
mail. Could that be right? If so, where do I start in trying to fix
it?
> That said, I do not know enough about system mail to point you to
> possible sources of your problem, however I would start by running mail
> with the verbose flag.

How do I do that?
> > Is this significant?
>
> I would say, definitely so.
>
> Hope this helps,

Every little bit of knowledge helps. Thanks!

Mark

[Suvayu Ali]

On Mon, Apr 14, 2014 at 08:03:55PM +0100, Arthur Dent wrote:
> On Mon, 2014-04-14 at 20:25 +0200, Suvayu Ali wrote:
> > > [mark@Zaphod]$ send-mail: account default not found: no configuration file available
> >
> > It looks like this is the crux of your problem.  I presume you checked
> > mary gets mails for her cron jobs?
>
> Well I think we're narrowing down the problem here. The answer to that
> question is NO!
> And interestingly NEITHER DOES ROOT!

This is strange.
> It seems that fixing /etc/aliases now allows me to get SYSTEM messages
> (e.g. logwatch, fail2ban stops/starts etc.) but NOT cron output.
>
> So the problem seems to be that cron is somehow not configured to send
> mail. Could that be right? If so, where do I start in trying to fix
> it?

I find this very surprising.  Can you try adding a line like this on top
of your crontab:

MAILTO=<username>
> > That said, I do not know enough about system mail to point you to
> > possible sources of your problem, however I would start by running mail
> > with the verbose flag.
>
> How do I do that?

Just call mail with -v in your test:

$ mail -v mary

Cheers,

F20 Where's my system mail? SOLVED (I Hope)

[Arthur Dent]

Hello all, With what (I hope) will be my final update on this issue.

This machine is a simple home server. It runs headless and is on 24/7. I
an in the habit running yum update once per month and only then
rebooting (and only then because the yum update usually brings down a
new kernel). I long ago left the Windows world where a reboot was
required for - well everything really. Since this is a new install
however I had rebooted a few more times than normal, but the last one
was a few days ago.

Last night (more in desperation than anything else) I rebooted.

Today - Cron emails!

I am very sorry for wasting everyone's time. I have no idea what the
problem was or why a reboot solved it - some environment variable
perhaps?

Anyway I am now a happy bunny and, as always, grateful for the friendly
and helpful support given here.

Thanks again.

Mark

F20 Where's my system mail? SOLVED (I Hope)

[Andrew Haley]

On 04/16/2014 09:31 AM, Arthur Dent wrote:
> Hello all, With what (I hope) will be my final update on this issue.
>
> This machine is a simple home server. It runs headless and is on 24/7. I an in the habit running yum update once per month and only then rebooting (and only then because the yum update usually brings down a new kernel). I long ago left the Windows world where a reboot was required for - well everything really. Since this is a new install however I had rebooted a few more times than normal, but the last one was a few days ago.
>
> Last night (more in desperation than anything else) I rebooted.
>
> Today - Cron emails!
>
> I am very sorry for wasting everyone's time. I have no idea what the problem was or why a reboot solved it - some environment variable perhaps?

Sendmail is like that.  This doesn't happen very often today, but in
the past people sometimes received messages that had been sent to them
months (if not years) before.  What had happened was that a mail
process had crashed leaving things in a bad state.  The server was
later rebooted, sendmail saw all the mails, and started to deliver
them.  It must have been distressing to receive an old message from an
estranged partner or someone who had died.

Andrew.

[poma]

On 10.04.2014 17:47, Arthur Dent wrote:
> OK - So if just do "yum install sendmail" will everything work as
> before? Or will I have to spend a fortune on coffee and pizzas as I sit
> up night after night trying to configure it?

What is tomorrow's weather forecast para las Islas Canarias?

poma