netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: David Laight <David.Laight@ACULAB.COM>,
	'Eric Dumazet' <edumazet@google.com>,
	Tal Gilboa <talgi@mellanox.com>
Cc: David Miller <davem@davemloft.net>,
	"ncardwell@google.com" <ncardwell@google.com>,
	"ycheng@google.com" <ycheng@google.com>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	Saeed Mahameed <saeedm@mellanox.com>,
	Tariq Toukan <tariqt@mellanox.com>,
	Amir Ancel <amira@mellanox.com>
Subject: Re: [PATCH net-next 0/7] tcp: implement rb-tree based retransmit queue
Date: Tue, 06 Feb 2018 07:52:18 -0800	[thread overview]
Message-ID: <1517932338.3715.149.camel@gmail.com> (raw)
In-Reply-To: <db881824dd3c4fc1b7b7b9e7f7489a80@AcuMS.aculab.com>

On Tue, 2018-02-06 at 15:22 +0000, David Laight wrote:
> From: Eric Dumazet
> > Sent: 06 February 2018 14:20
> 
> ...
> > Please give exact details.
> > Sending 64, 128, 256 or 512 bytes at a time on TCP_STREAM makes little sense.
> > We are not optimizing stack for pathological cases, sorry.
> 
> There are plenty of workloads which are not bulk data and where multiple
> small buffers get sent at unknown intervals (which may be back to back).
> Such connections have to have Nagle disabled because the Nagle delays
> are 'horrid'.
> Clearly lost packets can cause delays, but they are rare on local networks.

Auto corking makes sure aggregation happens, even for when Nagle is in
the picture.

netperf -- -m 256    will still cook 64KB TSO packets

netperf is not adding delays between each send(), unless it has been
modified.

  reply	other threads:[~2018-02-06 15:52 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-06  5:21 [PATCH net-next 0/7] tcp: implement rb-tree based retransmit queue Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 1/7] net: add rb_to_skb() and other rb tree helpers Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 2/7] tcp: uninline tcp_write_queue_purge() Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 3/7] tcp: tcp_tx_timestamp() cleanup Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 4/7] tcp: tcp_mark_head_lost() optimization Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 5/7] tcp: reduce tcp_fastretrans_alert() verbosity Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 6/7] tcp: pass previous skb to tcp_shifted_skb() Eric Dumazet
2017-10-06  5:21 ` [PATCH net-next 7/7] tcp: implement rb-tree based retransmit queue Eric Dumazet
2017-10-06 23:31 ` [PATCH net-next 0/7] " David Miller
2018-01-21 20:52   ` Tal Gilboa
2018-01-21 23:47     ` Eric Dumazet
2018-01-24 14:42       ` Tal Gilboa
2018-01-24 15:09         ` Eric Dumazet
2018-02-06 13:51           ` Tal Gilboa
2018-02-06 14:19             ` Eric Dumazet
2018-02-06 15:22               ` David Laight
2018-02-06 15:52                 ` Eric Dumazet [this message]
2018-02-06 16:27                   ` Tal Gilboa
2018-02-06 17:05                     ` Eric Dumazet

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=1517932338.3715.149.camel@gmail.com \
    --to=eric.dumazet@gmail.com \
    --cc=David.Laight@ACULAB.COM \
    --cc=amira@mellanox.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=ncardwell@google.com \
    --cc=netdev@vger.kernel.org \
    --cc=saeedm@mellanox.com \
    --cc=talgi@mellanox.com \
    --cc=tariqt@mellanox.com \
    --cc=ycheng@google.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).