From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jun'ichi Nomura Date: Wed, 01 Aug 2007 17:33:42 -0400 Subject: Re: [LVM2 PATCH 2/2] lvrename: add mirrored LV support In-Reply-To: <24987A0D-5C3A-4339-BCB8-F387BBC6A1A1@redhat.com> References: <46AA1024.5010606@ce.jp.nec.com> <24987A0D-5C3A-4339-BCB8-F387BBC6A1A1@redhat.com> Message-ID: <46B0FC36.8060604@ce.jp.nec.com> List-Id: To: lvm-devel@redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi Jon, Thanks for testing and reviewing. Jonathan Brassow wrote: >> - if ((lv->status & MIRRORED) || >> - (lv->status & MIRROR_LOG) || >> - (lv->status & MIRROR_IMAGE)) { >> - log_error("Mirrored LV, \"%s\" cannot be renamed: %s", >> - lv->name, strerror(ENOSYS)); >> - return 0; >> - } >> - > > I don't think you want to remove this wholesale. Testing reveals that > you are allowed to change a sub-lv name by itself, which is not > something we want, I don't think. Instead, leave the checks in for > MIRROR_LOG and MIRROR_IMAGE. You are right. I didn't intend to remove the limitation. I fixed it and modified 2 other parts: - Removed "Failed on " messages from _for_each_sublv(), as they have no useful information. Callback function is responsible to put necessary info about failure. Added a comment about it. - Changed _rename_sub_lv() to return success if it can't determine a new name for the sub LV. e.g. if "lv1" is a sub LV of "lv0", renaming "lv0" will success without changing "lv1". This situation won't happen with current LVM2 code. But if such LVs exist, I think expected behaviour is leaving them unchanged and renaming others as far as possible. Revised patch is attached. (and fixed subject to "2/2" as it should have been) Thanks, -- Jun'ichi Nomura, NEC Corporation of America -------------- next part -------------- A non-text attachment was scrubbed... Name: lvm2-lvrename-support-mirror.patch Type: text/x-patch Size: 4821 bytes Desc: not available URL: