oddity on ms start - Found word(s) check out in the Text body

[Glenn Steen]

On 28 October 2011 14:55, Rick Cooper <rcooper at dwford.com> wrote:
> Glenn Steen wrote:
>> On 28 October 2011 10:01, Glenn Steen <glenn.steen at gmail.com> wrote:
>>> I have one of those in the making too (noticed the diff in
>>> initscript naming ... MailScanner != mailscanner), so will
>>> check/compare with my prod systems.
>>>
>>> Cheers
>>> --
>>> -- Glenn
>>>
>>> Den 28 okt 2011 03:08 skrev "Jeremy McSpadden" <jeremy at fluxlabs.net>:
>>>>
>>>> The MS rpm was pulled down from the Baruwa repo. I will build it
>>>> out a few more times to determine if it is the repo or not.
>>>>
>>>>
>>
>> Well well, there it is... The init script supplied by the "baruwa repo
>> version" of mailscanner isn't the same as the one provided by
>> mailscanner.info. Just to be sure, I've DL'd the same version from
>> both, extracred (rpm2cpio mailsc*.rpm | cpio -i -d) both, and done
>> some looking and some diffs.
>> I'm not saying the baruwa one is bad, just that it is different and
>> hence might suffer from less exposure than the MailScanner one. I'm
>> sure the baruwa maintainer has a good reason for the change.
>> Question becomes... what else differs between the packages? I'll do
>> some further diffs and  let you know.
>>
>
> Just looking at the suspect init script I would go to the rh_status_q()
> function and comment out the "> /dev/null 2>&1" part after rh_status and see
> what is being output there. My guess is you will find it believes something
> is already running and exits before executing start() where as when you run
> restart it does not go through the rh_status_q() call and thus starts the
> daemons.
>
And that "simething" is probably postfix, which will be killed by a
restart action, perfectly explaining the behaviour.

Cheers
-- 
-- Glenn
email: glenn < dot > steen < at > gmail < dot > com
work: glenn < dot > steen < at > ap1 < dot > se