From: Sisu Xi <xisisu@gmail.com>
To: xen-devel@lists.xensource.com
Subject: network IO buffer? -- Sisu
Date: Wed, 7 Sep 2011 15:52:21 -0500 [thread overview]
Message-ID: <CAPqOm-rN6zjun+3h8wVmXE7c6=QkEACfUQh7h4zWoy7RWegPrg@mail.gmail.com> (raw)
Hi, all:
I want to study the latency in receiving a pkt in Xen.
Here is my set up: kernel version is 2.6.32.25, applied the patch.
Credit Based Scheduler is used. Dom0 is with 1 VCPU, pinned to core 0,
while Dom1 and Dom2 (one VCPU each) are both pinned to another
separate core.
Then I run a echo program in Dom1 to another server. If there is no
interference, the round trip time was like 250 microseconds.
But, if I run a iperf in Dom2 to generate the maximum interference,
the round trip time in Dom1 was like 40 ms, 160 times the original
case!
Does anyone know where is the pkt buffered? I already checked the
net_rx_action in netback.c in Dom0, the length of the queue is like 1
or 2 most of the time, which meas the pkt is buffered some where else?
Could anyone help me with this?
Thanks very much!
Best!
Yours,
Sisu
reply other threads:[~2011-09-07 20:52 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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='CAPqOm-rN6zjun+3h8wVmXE7c6=QkEACfUQh7h4zWoy7RWegPrg@mail.gmail.com' \
--to=xisisu@gmail.com \
--cc=xen-devel@lists.xensource.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).