From mboxrd@z Thu Jan 1 00:00:00 1970 From: Petr Rockai Date: Mon, 26 Mar 2012 14:06:46 +0200 Subject: [PATCH][just nitpicking :)] Make vgscan behaviour consistent with pvscan In-Reply-To: <4F6B1F13.5080103@redhat.com> (Peter Rajnoha's message of "Thu, 22 Mar 2012 13:46:11 +0100") References: <4F6B1F13.5080103@redhat.com> Message-ID: <87398vsh0p.fsf@mornfall.net> List-Id: To: lvm-devel@redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi, Peter Rajnoha writes: > Just this minor thing... Are you OK with this? > > If using "--cache" opt, we directly scan and update lvmetad if running. > If not using "--cache" opt, we get the info from lvmetad if running. yes, I think this is good for UI consistency. One thing to watch out for though is on-disk metadata inconsistencies. Before the patch, any "vgscan" would notice those, but afterwards, only "vgscan --cache" can be used to detect bad on-disk metadata. We could also want to extend pvscan --cache to report (and maybe fix?) any metadata inconsistencies it may find. [snip; patch looks OK] Petr -- id' Ash = Ash; id' Dust = Dust; id' _ = undefined