From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Guilherme G. Piccoli" Subject: Re: [PATCH] md/raid0: Fail BIOs if their underlying block device is gone Date: Mon, 29 Jul 2019 17:49:54 -0300 Message-ID: References: <20190729193359.11040-1-gpiccoli@canonical.com> <20190730011850.2f19e140@natsu> <053c88e1-06ec-0db1-de8f-68f63a3a1305@canonical.com> <20190730013655.229020ea@natsu> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20190730013655.229020ea@natsu> Content-Language: en-US List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Roman Mamedov Cc: linux-block@vger.kernel.org, Song Liu , NeilBrown , linux-raid@vger.kernel.org, dm-devel@redhat.com, jay.vosburgh@canonical.com List-Id: linux-raid.ids On 29/07/2019 17:36, Roman Mamedov wrote: > On Mon, 29 Jul 2019 17:27:15 -0300 > "Guilherme G. Piccoli" wrote: > >> Hi Roman, I don't think this is usual setup. I understand that there are >> RAID10 (also known as RAID 0+1) in which we can have like 4 devices, and >> they pair in 2 sets of two disks using stripping, then these sets are >> paired using mirroring. This is handled by raid10 driver however, so it >> won't suffer for this issue. >> >> I don't think it's common or even makes sense to back a raid1 with 2 >> pure raid0 devices. > > It might be not a usual setup, but it is a nice possibility that you get with > MD. If for the moment you don't have drives of the needed size, but have > smaller drives. E.g.: > > - had a 2x1TB RAID1; > - one disk fails; > - no 1TB disks at hand; > - but lots of 500GB disks; > - let's make a 2x500GB RAID0 and have that stand in for the missing 1TB > member for the time being; > > Or here's for a detailed rationale of a more permanent scenario: > https://louwrentius.com/building-a-raid-6-array-of-mixed-drives.html > Oh, that's nice to know, thanks for the clarification Roman. I wasn't aware this was more or less common. Anyway, I agree with you: in this case, it's a weak point of raid0 to be so slow to react in case of failures in one member. I hope this patch helps to alleviate the issue. Cheers, Guilherme