From mboxrd@z Thu Jan 1 00:00:00 1970 From: Xose Vazquez Perez Subject: Re: [PATCH] multipath-tools: replace multipath configuration output Date: Wed, 7 Dec 2016 16:44:22 +0100 Message-ID: References: <20161124224427.12609-1-xose.vazquez@gmail.com> <1480064411.23852.1.camel@suse.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1480064411.23852.1.camel@suse.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Martin Wilck , dm-devel@redhat.com, Christophe Varoqui List-Id: dm-devel.ids On 11/25/2016 10:00 AM, Martin Wilck wrote: > I am not sure I see the merit of these changes. If it's really > necessary to change the wording of the log messages which people have > got used to, the new ones should be really more self-explanatory than > the old ones. Some of them are a riddle, wrapped in a mystery, inside an enigma. For example "LUN setting". "setting: multipath.conf multipaths section" is, by far, a better description. > The common "setting: " formatting is nice but IMHO this alone doesn't > really justify overthrowing old habits. I really don't care the marker name, "setting:" "conf:" ... But a standard one should be used. > On Thu, 2016-11-24 at 23:44 +0100, Xose Vazquez Perez wrote: >> sysfs setting -> setting: kernel sysfs >> >> detected setting -> setting: array autodetected >> >> controller setting -> setting: array configuration > Is "array" really more understandable to users than "controller"? "controller", what controller? local hba? array controller? "array" is unequivocal. > These settings come from hwentries, so in reality they're either part > of the built-in hwtable or of the multipath.conf "devices" section, or > am I wrong here? Also from /sys fs, from the array or auto-negotiated kernel<->array. >> internal default -> setting: multipath internal >> >> overrides setting -> setting: multipath.conf overrides section >> >> LUN setting -> setting: multipath.conf multipaths section >> >> config file default >> config file setting -> setting: multipath.conf defaults/devices >> section > Have you double-checked if this is correct for all options? Yes. I don't do random patches. > It would be helpful for users to be able to distinguish which config file section > the option was taken from (defaults/devices/multipaths/overrides). The > same argument applies to the hardware enries - being able to > distinguish built in hardware defaults from config file "devices" > settings would be really helpful, but this patch doesn't provide that > functionality. This is a first step. Anyhow only "multipath.conf defaults/devices" could be split. overrides and multipaths sections already are independent as you can see in libmultipath/propsel.c