[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1477049817.7065.56.camel@edumazet-glaptop3.roam.corp.google.com>
Date: Fri, 21 Oct 2016 04:36:57 -0700
From: Eric Dumazet <eric.dumazet@...il.com>
To: Jiri Kosina <jikos@...nel.org>
Cc: Jamal Hadi Salim <jhs@...atatu.com>, Phil Sutter <phil@....cc>,
Cong Wang <xiyou.wangcong@...il.com>,
Daniel Borkmann <daniel@...earbox.net>,
"David S. Miller" <davem@...emloft.net>,
linux-kernel@...r.kernel.org, netdev@...r.kernel.org
Subject: Re: [PATCH] net: sched: make default fifo qdiscs appear in the dump
On Fri, 2016-10-21 at 10:45 +0200, Jiri Kosina wrote:
> The original reason [1] for having hidden qdiscs (potential scalability
> issues in qdisc_match_from_root() with single linked list in case of large
> amount of qdiscs) has been invalidated by 59cc1f61f0 ("net: sched: convert
> qdisc linked list to hashtable").
>
> This allows us for bringing more clarity and determinism into the dump by
> making default pfifo qdiscs visible.
>
> [1] http://lkml.kernel.org/r/1460732328.10638.74.camel@edumazet-glaptop3.roam.corp.google.com
>
> Signed-off-by: Jiri Kosina <jkosina@...e.cz>
> ---
Some of us are dealing with huge HTB hierarchies, so adding default fifo
in the dump will add more data pumped from the kernel.
BwE [1] for instance dumps qdisc/classes every 5 seconds.
I guess we'll need to not pull this patch in our kernels.
This reminds me of an HTB patch to avoid dumping rate estimates for HTB.
I will send it today.
[1]
http://static.googleusercontent.com/media/research.google.com/en//pubs/archive/43838.pdf
Powered by blists - more mailing lists