Re: MTD RAID

From: Dongsheng Yang <dongsheng.yang@easystack.cn>
To: Richard Weinberger <richard@nod.at>,
	Boris Brezillon <boris.brezillon@free-electrons.com>
Cc: starvik@axis.com, jesper.nilsson@axis.com,
	Dongsheng Yang <yangds.fnst@cn.fujitsu.com>,
	linux-cris-kernel@axis.com, shengyong1@huawei.com,
	Ard Biesheuvel <ard.biesheuvel@linaro.org>,
	David Woodhouse <dwmw2@infradead.org>,
	dmitry.torokhov@gmail.com, dooooongsheng.yang@gmail.com,
	jschultz@xes-inc.com, fabf@skynet.be, mtownsend1973@gmail.com,
	linux-mtd@lists.infradead.org,
	Colin King <colin.king@canonical.com>,
	asierra@xes-inc.com, Brian Norris <computersforpeace@gmail.com>,
	Dongsheng Yang <dongsheng081251@gmail.com>
Subject: Re: MTD RAID
Date: Fri, 19 Aug 2016 18:30:13 +0800	[thread overview]
Message-ID: <57B6DFB5.8000208@easystack.cn> (raw)
In-Reply-To: <a89263d4-9748-9437-e11d-62b5a6a54b7c@nod.at>

Hi Richard.

On 08/19/2016 05:47 PM, Richard Weinberger wrote:
> Yang,
>
> Sorry when I ask already answered questions, the mail thread
> grows faster than I can type. ;)
>
> On 19.08.2016 11:08, Dongsheng Yang wrote:
>> But how can we handle the multiple chips problem? Some drivers
>> are combining multiple chips to one single mtd device, what the
>> mtd_concat is doing.
> mtd_concat is a horrid and old hack which was used to combine very
> small flashes.
>
>>> One last question? What's the real goal of this MTD-RAID layer? If
>>> that's about addressing the MLC/TLC NAND reliability problems, I don't
>>> think it's such a good idea.
>> Oh, that's not the main problem I want to solve. RAID-1 is just a possible
>>   extension base on my RAID framework.
>>
>> This work is started for only RAID0, which is used to take the use of lots
>> of flash to improve performance. Then I refactored it to a MTD RAID
>> framework. Then we can implement other raid level for mtd.
>>
>> Example:
>>      In our production, there are 40+ chips attached on one pcie card.
>> Then we need to simulate all of them into one mtd device. At the same
>> time, we need to consider how to manage these chips. Finally we chose
>> a RAID0 mode for them. And got a great performance result.
>>
>> So, the multiple chips scenario is the original problem I want to solve. And
>> then I found I can refactor it for other RAID-levels.
> Are your chips NAND flash?
> Combining non-NAND is not that hard, mtd_concat does already.
> With NAND, as Boris pointed out, it will become complicated.

Yes, it's NAND, but I think this layer will be working for all MTD devices.
>
> So, what you propose as MTD RAID is not really RAID.
> RAID is about redundant disks. i.e. you can replace them upon failure.
> The problem you address is combining multiple chips.
> IOW an SSD emulated in software.

Yes, the original idea is from it, but I am wishing to implement a
real RAID for mtd now. Although we are only using RAID0 in
our production. I am glad to implement other RAID-levels in community.

And yes, sounds not easy for NAND devices. But I will to figure
out the solutions for these problems Boris pointed out.

BTW, what's the other idea to solve the reliability problem in MLC 
currently in community.

Yang
>
> I think we don't need replication on MTD. An improved mtd_concat
> does not sound bad but a proper implementation has to deal with
> all the nastiness of NAND.
>
> Thanks,
> //richard
>
> ______________________________________________________
> Linux MTD discussion mailing list
> http://lists.infradead.org/mailman/listinfo/linux-mtd/
>