All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chaitanya Kulkarni <Chaitanya.Kulkarni@wdc.com>
To: Max Gurtovoy <maxg@mellanox.com>,
	"sagi@grimberg.me" <sagi@grimberg.me>, "hch@lst.de" <hch@lst.de>
Cc: "linux-nvme@lists.infradead.org" <linux-nvme@lists.infradead.org>
Subject: Re: [RFC PATCH 1/2] nvmet: add bdev-ns polling support
Date: Thu, 30 Jan 2020 18:19:43 +0000	[thread overview]
Message-ID: <BYAPR04MB57492FCA245C3F19079CDD5186040@BYAPR04MB5749.namprd04.prod.outlook.com> (raw)
In-Reply-To: 03494a32-b31a-432b-870d-7347f5cc9596@mellanox.com

On 01/23/2020 06:23 AM, Max Gurtovoy wrote:
> On 1/21/2020 9:22 PM, Chaitanya Kulkarni wrote:
>> >Please see my in-line comments.
>> >On 01/20/2020 04:52 AM, Max Gurtovoy wrote:
>>> >>On 12/10/2019 8:25 AM, Chaitanya Kulkarni wrote:
>>>> >>>This patch adds support for the bdev-ns polling. We also add a new
>>>> >>>file to keep polling code separate (io-poll.c). The newly added
>>>> >>>configfs attribute allows user to enable/disable polling.
>>>> >>>
>>>> >>>We only enable polling for the READ/WRITE operation based on support
>>>> >>>from bdev and and use_poll configfs attr.
>>>> >>>
>>>> >>>We configure polling per namespace. For each namespace we create
>>>> >>>polling threads. For general approach please have a look at the
>>>> >>>cover-letter of this patch-series.
>>> >>It would be great to get some numbers here to learn about the trade-off
>>> >>for this approach.
>>> >>
>>> >>
>> >Already posted these numbers with QEMU nvme-loop here :-
>> >https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.infradead.org%2Fpipermail%2Flinux-nvme%2F2020-January%2F028692.html&amp;data=02%7C01%7Cmaxg%40mellanox.com%7C9308e8badf9f41ec4faf08d79ea733ae%7Ca652971c7d2e4d9ba6a4d149256f461b%7C0%7C0%7C637152313279111655&amp;sdata=RM3%2FcVeRcp8hgyAUFIM%2B7wbZ0cw2xvhdsfcVp9FuPm4%3D&amp;reserved=0
>> >
>> >IOPS/BW:-
>> >
>> >Default:-
>> >read: IOPS=52.8k, BW=206MiB/s (216MB/s)(6188MiB/30001msec)
>> >read: IOPS=54.3k, BW=212MiB/s (223MB/s)(6369MiB/30001msec)
>> >read: IOPS=53.5k, BW=209MiB/s (219MB/s)(6274MiB/30001msec)
>> >Poll:-
>> >read: IOPS=68.4k, BW=267MiB/s (280MB/s)(8011MiB/30001msec)
>> >read: IOPS=69.3k, BW=271MiB/s (284MB/s)(8124MiB/30001msec)
>> >read: IOPS=69.4k, BW=271MiB/s (284MB/s)(8132MiB/30001msec)
>> >
>> >mpstat:-
>> >Default:-
>> >CPU %usr %nice |%sys| %iowait %irq %soft %steal %guest %gnice %idle
>> >---------------------------------------------------------
>> >all 1.18 0.00 |60.14| 0.00 0.00 0.00 0.08 0.00 0.00 38.60
>> >---------------------------------------------------------
>> >0 0.00 0.00 | 32.00 |0.00 0.00 0.00 0.00 0.00 0.00 68.00
>> >1 1.01 0.00 | 42.42 |0.00 0.00 0.00 0.00 0.00 0.00 56.57
>> >2 1.01 0.00 | 57.58 |0.00 0.00 0.00 0.00 0.00 0.00 41.41
>> >3 2.02 0.00 | 79.80 |0.00 0.00 0.00 0.00 0.00 0.00 18.18
>> >4 1.01 0.00 | 63.64 |0.00 0.00 0.00 0.00 0.00 0.00 35.35
>> >5 3.09 0.00 | 63.92 |0.00 0.00 0.00 0.00 0.00 0.00 32.99
>> >6 2.02 0.00 | 75.76 |0.00 0.00 0.00 0.00 0.00 0.00 22.22
>> >7 1.02 0.00 | 57.14 |0.00 0.00 0.00 0.00 0.00 0.00 41.84
>> >8 0.00 0.00 | 67.01 |0.00 0.00 0.00 0.00 0.00 0.00 32.99
>> >9 1.01 0.00 | 59.60 |0.00 0.00 0.00 0.00 0.00 0.00 39.39
>> >10 1.02 0.00| 62.24 |0.00 0.00 0.00 0.00 0.00 0.00 36.73
>> >11 1.02 0.00| 62.24 |0.00 0.00 0.00 0.00 0.00 0.00 36.73
>> >
>> >Poll:-
>> >CPU %usr %nice |%sys| %iowait %irq %soft %steal %guest %gnice %idle
>> >---------------------------------------------------------
>> >all 1.08 0.00 98.08 0.00 0.00 0.00 0.08 0.00 0.00 0.75
>> >---------------------------------------------------------
>> >0 2.00 0.00 | 97.00 |0.00 0.00 0.00 0.00 0.00 0.00 1.00
>> >1 0.00 0.00 | 97.00 |0.00 0.00 0.00 0.00 0.00 0.00 3.00
>> >2 1.00 0.00 | 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >3 1.00 0.00 | 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >4 1.01 0.00 | 98.99 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >5 1.00 0.00 | 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >6 0.99 0.00 | 99.01 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >7 1.00 0.00 | 97.00 |0.00 0.00 0.00 0.00 0.00 0.00 2.00
>> >8 1.00 0.00 | 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >9 1.00 0.00 | 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
>> >10 1.98 0.00| 94.06 |0.00 0.00 0.00 0.00 0.00 0.00 3.96
>> >11 1.00 0.00| 99.00 |0.00 0.00 0.00 0.00 0.00 0.00 0.00
> with all the respect to nvmet-loop, it isn't a real use case, right ?
>
> Can you re-run it using RDMA and TCP ?
>
> what is the backing device for these tests ? NVMe/NULL_BLK ?
>
>
I don't think null_blk supports polling. This is NVMe QEMU Device.

Don't have the RDMA setup, I'll create one and report numbers.






_______________________________________________
linux-nvme mailing list
linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

  reply	other threads:[~2020-01-30 18:20 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-12-10  6:25 [RFC PATCH 0/2] nvmet: add polling support Chaitanya Kulkarni
2019-12-10  6:25 ` [RFC PATCH 1/2] nvmet: add bdev-ns " Chaitanya Kulkarni
2020-01-20 12:52   ` Max Gurtovoy
2020-01-21 19:22     ` Chaitanya Kulkarni
2020-01-23 14:23       ` Max Gurtovoy
2020-01-30 18:19         ` Chaitanya Kulkarni [this message]
2019-12-10  6:25 ` [RFC PATCH 2/2] nvmet: add file-ns " Chaitanya Kulkarni
2019-12-12  1:01 ` [RFC PATCH 0/2] nvmet: add " Sagi Grimberg
2019-12-12  5:44   ` Chaitanya Kulkarni
2019-12-12 20:32     ` Sagi Grimberg
2020-01-20  5:13       ` Chaitanya Kulkarni
2020-01-20  4:48   ` Chaitanya Kulkarni

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=BYAPR04MB57492FCA245C3F19079CDD5186040@BYAPR04MB5749.namprd04.prod.outlook.com \
    --to=chaitanya.kulkarni@wdc.com \
    --cc=hch@lst.de \
    --cc=linux-nvme@lists.infradead.org \
    --cc=maxg@mellanox.com \
    --cc=sagi@grimberg.me \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.