[katello-devel] katello-upgrade restarting services

[Eric Sammons]

----- Original Message -----
> On Tue 24 Jul 2012 07:36:22 AM MDT, Sam Kottler wrote:
> > How about checking if the service is running and then reacting
> > based on
> > that. If the service is running then katello-upgrade command will
> > exit and
> > alert the sysadmin that he/she should stop the service. If not,
> > then
> > upgrade can continue.
> >
> > Thoughts?
> >
> > On Tue, Jul 24, 2012 at 9:27 AM, Jordan OMara<jomara at redhat.com>
> >  wrote:
> >
> >> eric filed this bug:
> >> https://bugzilla.redhat.com/**show_bug.cgi?id=820280<https://bugzilla.redhat.com/show_bug.cgi?id=820280>
> >> og filed a similar one: https://bugzilla.redhat.com/**
> >> show_bug.cgi?id=837866<https://bugzilla.redhat.com/show_bug.cgi?id=837866>

I opened the bug because unfortunately SAM and CFSE have not just one rc script but many and the order may be a factor here, to protect the databases (I'm not sure).  I feel like, for this to feel like a finished product, that it should stop the services correctly.  As a safety measure though I do not think that a simple [Y/N] will suffice, it should be a [Yes/No] and it should be very clear what is about to happen, perhaps even a second confirmation to be sure.  

i.e. when you run rpm -Uhv on any thing you never ask the sysadmin to stop / start any daemons.  So to remain consistent with that expectation the script should handle this.  It also protects the companies that try to automate the upgrade task (and yes there are customers that will do that) and it also protects sysadmins that are in a hurry to get the job done.  Keep in mind, if we don't include stopping the services in the script the sysadmin will simply write a script that does it for them.  They seek to have the least amount of down time, and typing out long commands == downtime.

My $0.02.

-- Eric

> >> a pull request:
> >> https://github.com/Katello/**katello/pull/331<https://github.com/Katello/katello/pull/331>
> >>
> >> However, lzap&  msuchy have brought up that this might be poor
> >> behavior or very confusing to a sysadmin and that they would want
> >> to
> >> almost always stop services themselves before starting the upgrade
> >> process. I don't have enough sysadmin experience to make a good
> >> decision so I wanted to bring it up to the list.
> >>
> >> tl;dr: katello-upgrade currently fails if you haven't stopped your
> >> services. the pull requests makes the script stop them for you
> >>
> >> my uninformed 2c: the patch is fairly innocuous because you can
> >> decline to have the script stop anything and exit
> >>
> >> DISCUSS
> >> --
> >> Jordan O'Mara<jomara at redhat.com>
> >> Red Hat Engineering, Raleigh
> >> _______________________________________________
> >> katello-devel mailing list
> >> katello-devel at redhat.com
> >> https://www.redhat.com/mailman/listinfo/katello-devel
> >>
> >>
> >
> >
> 
> 
> Seems obvious when you put it that way...
>