From: Anton Danilov <littlesmilingcloud@gmail.com>
To: lartc@vger.kernel.org
Subject: Re: Performance for unclassified packets when using HTB
Date: Tue, 04 Sep 2018 09:21:08 +0000 [thread overview]
Message-ID: <CAEzD07JcXMv966eYuq20WFNaRBXK2MhQNPCcRr8_xKFW0MR_=w@mail.gmail.com> (raw)
Hello.
The unclassified packets are being placed into X:0 direct class if
there isn't default class option of the htb discipline.
To avoid it specify the default class at the root qdisc creating.
On Fri, 10 Aug 2018 at 04:31, Abhishek Modi <abshkmodi@gmail.com> wrote:
>
> Hey,
>
> We are using HTB for a project and observing some performance
> degradation for network transfers. I could not find any documentation
> anywhere specifying what happens to unclassified packets - is there
> some rate limit on them due to filtering overhead or the addition of
> filters should not affect the transfer rate of unclassified packets ?
>
> This is what our configuration looks like -
>
> sudo tc qdisc add dev bond0 root handle 1: htb
> sudo tc class add dev bond0 parent 1: classid 1:10 htb rate
> 34359738360 burst 34359738360 cburst 34359738360
> sudo tc class add dev bond0 parent 1: classid 1:20 htb rate
> 34359738360 burst 34359738360 cburst 34359738360
> sudo tc filter add dev bond0 parent 1: prio 99 protocol ip u32 match
> ip dport 7700 0xffff classid 1:10
> sudo tc filter add dev bond0 parent 1: prio 100 protocol ip u32 match
> ip sport 7701 0xffff classid 1:20
>
> The limits for the classes are dynamic and we update them periodically
> based on the configuration. After adding these filters and classes, we
> are seeing 50% performance drop for traffic not going to/from any of
> the above ports (7700 and 7701).
>
> Is there anything wrong with our configuration ?
--
Anton.
next reply other threads:[~2018-09-04 9:21 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-04 9:21 Anton Danilov [this message]
-- strict thread matches above, loose matches on Subject: below --
2018-08-10 1:29 Performance for unclassified packets when using HTB Abhishek Modi
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='CAEzD07JcXMv966eYuq20WFNaRBXK2MhQNPCcRr8_xKFW0MR_=w@mail.gmail.com' \
--to=littlesmilingcloud@gmail.com \
--cc=lartc@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 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 read-only IMAP folder(s) and NNTP newsgroup(s).