All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jelle de Jong <jelledejong@powercraft.nl>
To: linux-ppp@vger.kernel.org
Subject: Re: slow transfer speed issue between two ppp based servers
Date: Fri, 15 Jan 2010 14:20:55 +0000	[thread overview]
Message-ID: <4B5079C7.9030706@powercraft.nl> (raw)
In-Reply-To: <4B4AFA68.4020500@powercraft.nl>

[-- Attachment #1: Type: text/plain, Size: 1539 bytes --]

James Carlson wrote, on 11-01-10 15:38:
> Is there any chance you could make the raw tcpdump files (rather than
> the decoded text output) available?  It's much harder to debug things
> when information has been subtracted.
> 
> Based on what little is available here, I'm seeing what seems to be an
> RTT that jumps around a lot, from a few milliseconds on up to 150
> milliseconds, and causing TCP to misbehave.  It's likely that this is a
> kernel issue that needs to be investigated.

I created an new tcpdump of the issue and attached it in the mail.

http://filebin.ca/zmmnws/tcpdump01.tar.gz

I changed the sammy servers pppoe connection to an normal eth5
interface to an ciso router that is providing a 50/50mbit fiber
interface. So the upload speed of the sammy server should be very
high. But is only 30KB when uploading to my mobile umts systems. When
uploading form an other server to the umts systems the upload speed is
the max of that server (100kB) so the umts systems should not be the
problem here.

We opened the tcmpdumps with wireshark and did an analyse > export
info and it showed a lot of "TCP Fast retransmissions" compared to the
the other tcmpdump of the good download from the ashley server.

I also tried the below kernels, but it did not change anything.

linux-image-686 | 2.6.30
linux-image-686 | 2.6.31
linux-image-686 | 2.6.32
linux-image-amd64 | 2.6.30
linux-image-amd64 | 2.6.32

If anybody can have a look at the tcp dumps or maybe have other
suggestion to get more debug info?

Best regards,

Jelle



[-- Attachment #2: setup-sammy-uplinkspeed-debug.txt --]
[-- Type: text/plain, Size: 2882 bytes --]

���������------------------------------------------------------------------------

# this document describes my issue with a way to slow transfer speed
# between my sammy server with pppoe g.shdsl and an alix05 server with
# ppp hsdpa umts

------------------------------------------------------------------------

# two download tests, demonstrating that the alix05 server can download
# with 250KB/s (umts hsdpa 3g)
alix05:~# wget -cr http://ftp.nl.debian.org/debian/ls-lR.gz
FINISHED --2010-01-15 13:43:41--
Downloaded: 1 files, 6.5M in 39s (168 KB/s)
FINISHED --2010-01-15 13:44:54--
Downloaded: 1 files, 6.5M in 46s (143 KB/s)


# one upload test to the sammy server, demonstrating that the alix05
# server can upload with 60KB/s and that the sammy server can download
# with at least 60KB's (umts hsdpa 3g)
alix05:~# scp ftp.nl.debian.org/debian/ls-lR.gz root@sammy.somewhere.nl:/root/
100% 6607KB  61.2KB/s   01:48

# one upload from the sammy server to the ashley server, demonstrating
# that the sammy server can upload with 274KB/s. (g.shdsl)
sammy:~# scp /root/ls-lR.gz root@powercraft.nl:/root/
100% 6608KB   1.3MB/s   00:05

### THIS IS THE ISSUE, THE DOWNLOAD SPEED OF 17KB/s IS WAY TO LOW ###
# one download from sammy server to alix05 server that is going to slow
alix05:~# scp root@sammy.somewhere.nl:/root/ls-lR.gz ls-lR.gz
100% 6578KB  17.4KB/s   06:19

# control test, downloading from the ashley server to demonstrate that
# the alix05 is capable of download at least with 97KB/s with scp
alix05:~# scp root@powercraft.nl:/root/ls-lR.gz ls-lR.gz
100% 6608KB  88.1KB/s   01:15

------------------------------------------------------------------------

root@ashley:~# tcpdump port 22 and host 62.140.137.47 -i eth1 -w tcpdump-raw-ashley-eth1.txt
    from
alix05:~# scp root@powercraft.nl:/root/ls-lR.gz ls-lR.gz
    100% 6608KB  89.3KB/s   01:14

------------------------------------------------------------------------

sammy:~# tcpdump port 22 and host 62.140.137.47 -i eth5 -w tcpdump-raw-sammy-eth5.txt
    from
alix05:~# scp root@sammy.somewhere.nl:/root/ls-lR.gz ls-lR.g
    100% 6608KB  24.5KB/s   04:30

------------------------------------------------------------------------

# othertests:
# scp root@10.10.1.230:/root/ls-lR.gz ls-lR.g

------------------------------------------------------------------------

mv /root/tcpdump-ashley-eth1.txt /home/jelle/tcpdump-raw-ashley-eth1.txt
chown jelle:jelle /home/jelle/tcpdump-raw-ashley-eth1.txt
chmod 644 /home/jelle/tcpdump-raw-ashley-eth1.txt

scp --verbose /root/tcpdump-raw-sammy-eth5.txt root@powercraft.nl:/root/
# on ashley
mv /root/tcpdump-raw-sammy-eth5.txt /home/jelle/tcpdump-raw-sammy-eth5.txt
chown jelle:jelle /home/jelle/tcpdump-raw-sammy-eth5.txt
chmod 644 /home/jelle/tcpdump-raw-sammy-eth5.txt

------------------------------------------------------------------------



  parent reply	other threads:[~2010-01-15 14:20 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-01-11 10:16 slow transfer speed issue between two ppp based servers Jelle de Jong
2010-01-11 14:38 ` James Carlson
2010-01-15 14:20 ` Jelle de Jong [this message]
2010-01-16 20:17 ` Jelle de Jong

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=4B5079C7.9030706@powercraft.nl \
    --to=jelledejong@powercraft.nl \
    --cc=linux-ppp@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.