[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <9679B268-5BD7-4731-86CC-6B5CF7CE2DF2@smartx.com>
Date: Thu, 28 Dec 2023 15:25:10 +0800
From: Li Feng <fengli@...rtx.com>
To: "Michael S. Tsirkin" <mst@...hat.com>
Cc: Paolo Bonzini <pbonzini@...hat.com>,
Jens Axboe <axboe@...nel.dk>,
Jason Wang <jasowang@...hat.com>,
Stefan Hajnoczi <stefanha@...hat.com>,
Xuan Zhuo <xuanzhuo@...ux.alibaba.com>,
"open list:BLOCK LAYER" <linux-block@...r.kernel.org>,
linux-kernel <linux-kernel@...r.kernel.org>,
"open list:VIRTIO BLOCK AND SCSI DRIVERS" <virtualization@...ts.linux.dev>
Subject: Re: [PATCH] virtio_blk: set the default scheduler to none
> On Dec 27, 2023, at 18:06, Michael S. Tsirkin <mst@...hat.com> wrote:
>
> On Wed, Dec 27, 2023 at 03:26:30PM +0800, Li Feng wrote:
>>
>>
>>> On Dec 26, 2023, at 23:38, Michael S. Tsirkin <mst@...hat.com> wrote:
>>>
>>> On Tue, Dec 26, 2023 at 05:01:40PM +0800, Li Feng wrote:
>>>> I don't have ideas right now to answer Christoph/Paolo's question.
>>>
>>> Paolo did some testing on this thread and posted some concerns.
>>> Do you disagree with his analysis?
>>
>> Paolo gave a very detailed data analysis. Indeed, in some low queue depth
>> scenarios, the data dropped somewhat. However, I suspect that it may be
>> caused by other problems (such as test fluctuations) rather than the benefits
>> brought by deadline.
>
> Maybe. I think it would be up to you to prove this then.
Ok.
>
>
>> BTW, I think 128 queue depth test is a very important and common performance
>> test item for storage devices.
>>
>> Thanks,
>> Li
>>
>>>
>>> --
>>> MST
>>>
>
Powered by blists - more mailing lists