[OmniOS-discuss] Postfix service: syslog reports "fatal: the Postfix mail system is already running" every second

[Andries Annema]

Thanks, Jim. That is indeed the suggested cause that I find all over the 
web, however even if I stop the postfix-service, then manually delete 
the .pid and .lock file and then restart the service, it still reports 
that it's already running.
I don't see how it could be.
Also, the very same package worked flawlessly on r151014, but now on 
r151022 it starts bugging me.



On 2017-07-30 22:00, Jim Klimov wrote:
> On July 30, 2017 5:42:31 PM GMT+02:00, Andries Annema <an3s.annema at gmail.com> wrote:
>> In the mean time I tried upgrading the separate test-system from
>> r151022
>> "vanilla" to the CE-edition -  r151022i by now - hoping this might
>> somehow also solve the below postfix "already running" error. But, it
>> didn't.
>>
>> Again I tried the solution that worked on r151014 and also a reboot,
>> but
>> without success.
>>
>> This "Postfix mail system is already running"-issue is all over the
>> internet - not specifically on Solarish-systems, that is - but I am
>> clearly missing something here, as I don't seem to get the issue out of
>>
>> there.
>>
>> Is there anyone out there with some ideas on this, pretty please?
>>
>> Thanks.
>>
>> Andries
>>
>>
>> On 2017-07-14 14:45, Andries Annema wrote:
>>> Hi all,
>>>
>>> Have been running postfix from the "ms.omniti.com" publisher for some
>>> years now on r151014 without issues.
>>> Last week I updated my October 2016 r151014 system to the latest
>>> r151014-status, as a first step in the direction of upgrading to the
>>> latest LTS. I'm not actually there yet to upgrade to r151022; first I
>>> want to make sure all my services and packages run flawlessly on that
>>> version. Turns out that being careful pays off.
>>>
>>> In this update process, postfix got updated to
>>> 2.11.9-0.151014:20170214T224831Z.
>>> It still works, but /var/log/syslog is filled with a new line ever
>>> second stating:
>>>
>>> ".... [ID 947731 mail.crit] fatal: the Postfix mail system is already
>>> running"
>>>
>>> And /var/svc/log/network-smtp-postfix:default.log is repeatedly
>> filled
>>> with line pairs like:
>>> "[ Jul 14 13:48:50 Executing start method ("/opt/omni/sbin/postfix
>>> start"). ]
>>> [ Jul 14 13:48:51 Stopping because service exited with an error. ]
>>> [ Jul 14 13:48:51 Executing start method ("/opt/omni/sbin/postfix
>>> start"). ]
>>> [ Jul 14 13:48:51 Stopping because service exited with an error. ]
>>> [ Jul 14 13:48:51 Failing too quickly, throttling. ]"
>>>
>>> Simply disabling and re-enabling the SMF service does not help.
>>> What did help was this:
>>> $ sudo svcadm disable postfix
>>> $ sudo fuser -k /var/lib/postfix/master.lock
>>> $ sudo rm /var/lib/postfix/master.lock
>>> $ sudo rm /var/spool/postfix/pid/master.pid
>>> $ sudo svcadm enable postfix
>>>
>>> That, however, was on r151014.
>>> On a separate, freshly installed r151022 test-system, I installed
>> this
>>> same postfix 2.11.9 package and the same issue emerged.
>>> However, the fix that worked on r151014, does not seem to work on
>>> r151022.
>>> Rebooting doesn't help either.
>>>
>>> Funny thing is: it does deliver email nicely, despite this restarting
>>> issue.
>>>
>>> I tried downgrading to postfix 2.10.10: problem still exists.
>>> Tried downgrading further, to 2.10.2: not possible, because the
>>> manifest states a "incorporate" dependency on the
>> "entire at 11-0.151014"
>>> package. Hm, right... now I'm out of ideas.
>>> Any clues?
>>>
>>> Regards,
>>> Andries
>>>
>> _______________________________________________
>> OmniOS-discuss mailing list
>> OmniOS-discuss at lists.omniti.com
>> http://lists.omniti.com/mailman/listinfo/omnios-discuss
> IIRC the root problem was that some pid-file was left in place before restarting, e.g. by ungraceful reboot or timed-out stop with a kill.
> --
> Typos courtesy of K-9 Mail on my Android