All of lore.kernel.org
 help / color / mirror / Atom feed
From: Stefan Hajnoczi <stefanha@redhat.com>
To: Jens Axboe <axboe@kernel.dk>
Cc: linux-kernel@vger.kernel.org, netdev@vger.kernel.org
Subject: Re: [PATCHSET v3 0/5] Add support for epoll min_wait
Date: Tue, 8 Nov 2022 09:00:49 -0500	[thread overview]
Message-ID: <Y2phEZKYuSmPL5B5@fedora> (raw)
In-Reply-To: <4281b354-d67d-2883-d966-a7816ed4f811@kernel.dk>

[-- Attachment #1: Type: text/plain, Size: 1543 bytes --]

On Mon, Nov 07, 2022 at 02:38:52PM -0700, Jens Axboe wrote:
> On 11/7/22 1:56 PM, Stefan Hajnoczi wrote:
> > Hi Jens,
> > NICs and storage controllers have interrupt mitigation/coalescing
> > mechanisms that are similar.
> 
> Yep
> 
> > NVMe has an Aggregation Time (timeout) and an Aggregation Threshold
> > (counter) value. When a completion occurs, the device waits until the
> > timeout or until the completion counter value is reached.
> > 
> > If I've read the code correctly, min_wait is computed at the beginning
> > of epoll_wait(2). NVMe's Aggregation Time is computed from the first
> > completion.
> > 
> > It makes me wonder which approach is more useful for applications. With
> > the Aggregation Time approach applications can control how much extra
> > latency is added. What do you think about that approach?
> 
> We only tested the current approach, which is time noted from entry, not
> from when the first event arrives. I suspect the nvme approach is better
> suited to the hw side, the epoll timeout helps ensure that we batch
> within xx usec rather than xx usec + whatever the delay until the first
> one arrives. Which is why it's handled that way currently. That gives
> you a fixed batch latency.

min_wait is fine when the goal is just maximizing throughput without any
latency targets.

The min_wait approach makes it hard to set a useful upper bound on
latency because unlucky requests that complete early experience much
more latency than requests that complete later.

Stefan

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

  reply	other threads:[~2022-11-08 14:02 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-30 22:01 [PATCHSET v3 0/5] Add support for epoll min_wait Jens Axboe
2022-10-30 22:01 ` [PATCH 1/6] eventpoll: cleanup branches around sleeping for events Jens Axboe
2022-10-30 22:01 ` [PATCH 2/6] eventpoll: don't pass in 'timed_out' to ep_busy_loop() Jens Axboe
2022-10-30 22:02 ` [PATCH 3/6] eventpoll: split out wait handling Jens Axboe
2022-10-30 22:02 ` [PATCH 4/6] eventpoll: move expires to epoll_wq Jens Axboe
2022-10-30 22:02 ` [PATCH 5/6] eventpoll: move file checking earlier for epoll_ctl() Jens Axboe
2022-10-30 22:02 ` [PATCH 6/6] eventpoll: add support for min-wait Jens Axboe
2022-11-08 22:14   ` Soheil Hassas Yeganeh
2022-11-08 22:20     ` Jens Axboe
2022-11-08 22:25       ` Willem de Bruijn
2022-11-08 22:29         ` Jens Axboe
2022-11-08 22:44           ` Willem de Bruijn
2022-11-08 22:41       ` Soheil Hassas Yeganeh
2022-12-01 18:00       ` Jens Axboe
2022-12-01 18:39         ` Soheil Hassas Yeganeh
2022-12-01 18:41           ` Jens Axboe
2022-11-02 17:46 ` [PATCHSET v3 0/5] Add support for epoll min_wait Willem de Bruijn
2022-11-02 17:54   ` Jens Axboe
2022-11-02 23:09     ` Willem de Bruijn
2022-11-02 23:37       ` Jens Axboe
2022-11-02 23:51         ` Willem de Bruijn
2022-11-02 23:57           ` Jens Axboe
2022-11-05 17:39             ` Jens Axboe
2022-11-05 18:05               ` Willem de Bruijn
2022-11-05 18:46                 ` Jens Axboe
2022-11-07 13:25                   ` Willem de Bruijn
2022-11-07 14:19                     ` Jens Axboe
2022-11-07 10:10               ` David Laight
2022-11-07 20:56 ` Stefan Hajnoczi
2022-11-07 21:38   ` Jens Axboe
2022-11-08 14:00     ` Stefan Hajnoczi [this message]
2022-11-08 14:09       ` Jens Axboe
2022-11-08 16:10         ` Stefan Hajnoczi
2022-11-08 16:15           ` Jens Axboe
2022-11-08 17:24             ` Stefan Hajnoczi
2022-11-08 17:28               ` Jens Axboe
2022-11-08 20:29                 ` Stefan Hajnoczi
2022-11-09 10:09               ` David Laight
2022-11-10 10:13         ` Willem de Bruijn

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=Y2phEZKYuSmPL5B5@fedora \
    --to=stefanha@redhat.com \
    --cc=axboe@kernel.dk \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    /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.