Re: [tcpm] TCP tuning

To: rick jones <perfgeek at mac.com>
Subject: Re: [tcpm] TCP tuning
From: rick jones <perfgeek at mac.com>
Date: Wed, 03 Feb 2010 22:09:00 -0800
Cc: tcpm at ietf.org, "SCHARF, Michael" <Michael.Scharf at alcatel-lucent.com>, Joe Touch <touch at ISI.EDU>
Delivered-to: tcpm at core3.amsl.com
In-reply-to: <9E43CA53-5BFD-4377-93D1-1AED32C09B0E at mac.com>
List-archive: <http://www.ietf.org/mail-archive/web/tcpm>
List-help: <mailto:tcpm-request@ietf.org?subject=help>
List-id: TCP Maintenance and Minor Extensions Working Group <tcpm.ietf.org>
List-post: <mailto:tcpm@ietf.org>
List-subscribe: <https://www.ietf.org/mailman/listinfo/tcpm>, <mailto:tcpm-request@ietf.org?subject=subscribe>
List-unsubscribe: <https://www.ietf.org/mailman/listinfo/tcpm>, <mailto:tcpm-request@ietf.org?subject=unsubscribe>
References: <7BE9742D-6EDC-43FE-84FC-D22C52D23152 at nokia.com> <133D9897FB9C5E4E9DF2779DC91E947C025F1861 at SLFSNX.rcs.alcatel-research.de> <d1c2719f1002031110v3b76ca9eu14c9a110847548e7 at mail.gmail.com> <4B69CDD7.6060802 at isi.edu> <E086E248-DEFA-4AA4-B25D-F7FBB0FB9E7D at mac.com> <4B6A43D1.2010706 at isi.edu> <9E43CA53-5BFD-4377-93D1-1AED32C09B0E at mac.com>

On Feb 3, 2010, at 9:41 PM, rick jones wrote:

PS - I suppose if I'd been thinking more clearly, I'd have alsoincluded --enable-historgram in the ./configure command and beenable to give a histogram of the response times (adding a -v 2 optionto the global command line options), but I suppose that can beguesstimated from the traces.

Ok, so perhaps I like running netperf too much :) Here then are theapplication-level measured response times. The first netperf is withtcp_slow_start_after_idle = 0, the second with it set to 1, I'll leave-out the rest of the netperf output as it is repetitive:


tcp_slow_start_after_idle = 0 on server:

Histogram of request/response times

UNIT_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0HUNDRED_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0UNIT_MSEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_MSEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0HUNDRED_MSEC : 0: 294: 4: 0: 2: 0: 0: 0:0: 0UNIT_SEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_SEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0

>100_SECS: 0
HIST_TOTAL:      300

That means 294 times 100ms <= responsetime < 200ms, 4 times it was200ms <= responsetime < 300ms etc etc.


tcp_slow_start_after_idle = 1 on server:

UNIT_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0HUNDRED_USEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0UNIT_MSEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_MSEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0HUNDRED_MSEC : 0: 0: 0: 290: 10: 0: 0: 0:0: 0UNIT_SEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0TEN_SEC : 0: 0: 0: 0: 0: 0: 0: 0:0: 0

>100_SECS: 0
HIST_TOTAL:      300

Now it was 290 with 300ms <= responsetime < 400ms

Just as a refresher, the request size was 128 bytes, the response size16384.


The ICMP Echo RTT is ~120 ms between my home system and netperf.org:

--- netperf.org ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max/stddev = 119.523/120.410/121.248/0.493 ms

rick jones
Wisdom teeth are impacted, people are affected by the effects of events

Follow-Ups:
- Re: [tcpm] TCP tuning
  - From: Joe Touch

References:
- [tcpm] TCP tuning
  - From: Lars Eggert
- Re: [tcpm] TCP tuning
  - From: SCHARF, Michael
- Re: [tcpm] TCP tuning
  - From: Jerry Chu
- Re: [tcpm] TCP tuning
  - From: Joe Touch
- Re: [tcpm] TCP tuning
  - From: rick jones
- Re: [tcpm] TCP tuning
  - From: Joe Touch
- Re: [tcpm] TCP tuning
  - From: rick jones

Prev by Date: Re: [tcpm] TCP tuning
Next by Date: Re: [tcpm] TCP tuning
Previous by thread: Re: [tcpm] TCP tuning
Next by thread: Re: [tcpm] TCP tuning
Index(es):
- Date
- Thread

Re: [tcpm] TCP tuning

Note: Messages sent to this list are the opinions of the senders and do not imply endorsement by the IETF.