From: ebiederm@xmission•com (Eric W. Biederman)
To: Eric Dumazet <edumazet@google•com>
Cc: netdev <netdev@vger•kernel.org>,
Stephen Hemminger <stephen@networkplumber•org>,
Nicolas Dichtel <nicolas.dichtel@6wind•com>,
roopa <roopa@cumulusnetworks•com>,
Hannes Frederic Sowa <hannes@stressinduktion•org>,
Dinesh Dutt <ddutt@cumulusnetworks•com>,
Vipin Kumar <vipin@cumulusnetworks•com>,
Shmulik Ladkani <shmulik.ladkani@gmail•com>,
David Ahern <dsahern@gmail•com>,
"David S. Miller" <davem@davemloft•net>
Subject: Re: network namespace bloat
Date: Tue, 10 Feb 2015 21:18:13 -0600 [thread overview]
Message-ID: <87d25hrv9m.fsf@x220.int.ebiederm.org> (raw)
In-Reply-To: <CANn89iJ2auNoChr3FuFh-AGaPaw5_wwqfjZrtup1+xu6reVHMA@mail.gmail.com> (Eric Dumazet's message of "Tue, 10 Feb 2015 18:55:39 -0800")
Eric Dumazet <edumazet@google•com> writes:
> On Tue, Feb 10, 2015 at 6:42 PM, Eric W. Biederman
> <ebiederm@xmission•com> wrote:
>
>>
>> Those large hash tables impact creation speed as large memory
>> allocations require more work from the memory allocators, and they
>> affect reliability of as order > 0 pages are not reliabily available in
>> the kernel. So from a network namespace perspective I would really like
>> to convert the per network namespace hash tables to hash tables that
>> have a single instance across all network namespaces.
>>
>
> tcp_metric can fallback to vzalloc() after commit 976a702ac9eea ?
True, although vmalloc space is limited on some architectures.
> There is nothing preventing to use a single tcp_metric, a bit like
> global TCP hash table.
>
> We only have to convert the thing...
Thanks for the confirmation, that is what I figured was going on.
> Note : At Google we do not save tcp metrics.
> We have to use it only for FastOpen cookies eventually (For clients)
Interesting. That makes it doubly desirable to not need to allocate
a per network namespace hash table.
Eric
next prev parent reply other threads:[~2015-02-11 3:21 UTC|newest]
Thread overview: 119+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-02-05 1:34 [RFC PATCH 00/29] net: VRF support David Ahern
2015-02-05 1:34 ` [RFC PATCH 01/29] net: Introduce net_ctx and macro for context comparison David Ahern
2015-02-05 1:34 ` [RFC PATCH 02/29] net: Flip net_device to use net_ctx David Ahern
2015-02-05 13:47 ` Nicolas Dichtel
2015-02-06 0:45 ` David Ahern
2015-02-05 1:34 ` [RFC PATCH 03/29] net: Flip sock_common to net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 04/29] net: Add net_ctx macros for skbuffs David Ahern
2015-02-05 1:34 ` [RFC PATCH 05/29] net: Flip seq_net_private to net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 06/29] net: Flip fib_rules and fib_rules_ops to use net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 07/29] net: Flip inet_bind_bucket to net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 08/29] net: Flip fib_info " David Ahern
2015-02-05 1:34 ` [RFC PATCH 09/29] net: Flip ip6_flowlabel " David Ahern
2015-02-05 1:34 ` [RFC PATCH 10/29] net: Flip neigh structs " David Ahern
2015-02-05 1:34 ` [RFC PATCH 11/29] net: Flip nl_info " David Ahern
2015-02-05 1:34 ` [RFC PATCH 12/29] net: Add device lookups by net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 13/29] net: Convert function arg from struct net to struct net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 14/29] net: vrf: Introduce vrf header file David Ahern
2015-02-05 13:44 ` Nicolas Dichtel
2015-02-06 0:52 ` David Ahern
2015-02-06 8:53 ` Nicolas Dichtel
2015-02-05 1:34 ` [RFC PATCH 15/29] net: vrf: Add vrf to net_ctx struct David Ahern
2015-02-05 1:34 ` [RFC PATCH 16/29] net: vrf: Set default vrf David Ahern
2015-02-05 1:34 ` [RFC PATCH 17/29] net: vrf: Add vrf context to task struct David Ahern
2015-02-05 1:34 ` [RFC PATCH 18/29] net: vrf: Plumbing for vrf context on a socket David Ahern
2015-02-05 13:44 ` Nicolas Dichtel
2015-02-06 1:18 ` David Ahern
2015-02-05 1:34 ` [RFC PATCH 19/29] net: vrf: Add vrf context to skb David Ahern
2015-02-05 13:45 ` Nicolas Dichtel
2015-02-06 1:21 ` David Ahern
2015-02-06 3:54 ` Eric W. Biederman
2015-02-06 6:00 ` David Ahern
2015-02-05 1:34 ` [RFC PATCH 20/29] net: vrf: Add vrf context to flow struct David Ahern
2015-02-05 1:34 ` [RFC PATCH 21/29] net: vrf: Add vrf context to genid's David Ahern
2015-02-05 1:34 ` [RFC PATCH 22/29] net: vrf: Set VRF id in various network structs David Ahern
2015-02-05 1:34 ` [RFC PATCH 23/29] net: vrf: Enable vrf checks David Ahern
2015-02-05 1:34 ` [RFC PATCH 24/29] net: vrf: Add support to get/set vrf context on a device David Ahern
2015-02-05 1:34 ` [RFC PATCH 25/29] net: vrf: Handle VRF any context David Ahern
2015-02-05 13:46 ` Nicolas Dichtel
2015-02-06 1:23 ` David Ahern
2015-02-05 1:34 ` [RFC PATCH 26/29] net: vrf: Change single_open_net to pass net_ctx David Ahern
2015-02-05 1:34 ` [RFC PATCH 27/29] net: vrf: Add vrf checks and context to ipv4 proc files David Ahern
2015-02-05 1:34 ` [RFC PATCH 28/29] iproute2: vrf: Add vrf subcommand David Ahern
2015-02-05 1:34 ` [RFC PATCH 29/29] iproute2: Add vrf option to ip link command David Ahern
2015-02-05 5:17 ` [RFC PATCH 00/29] net: VRF support roopa
2015-02-05 13:44 ` Nicolas Dichtel
2015-02-06 1:32 ` David Ahern
2015-02-06 8:53 ` Nicolas Dichtel
2015-02-05 23:12 ` roopa
2015-02-06 2:19 ` David Ahern
2015-02-09 16:38 ` roopa
2015-02-10 10:43 ` Derek Fawcus
2015-02-06 6:10 ` Shmulik Ladkani
2015-02-09 15:54 ` roopa
2015-02-11 7:42 ` Shmulik Ladkani
2015-02-06 1:33 ` Stephen Hemminger
2015-02-06 2:10 ` David Ahern
2015-02-06 4:14 ` Eric W. Biederman
2015-02-06 6:15 ` David Ahern
2015-02-06 15:08 ` Nicolas Dichtel
[not found] ` <87iofe7n1x.fsf@x220.int.ebiederm.org>
2015-02-09 20:48 ` Nicolas Dichtel
2015-02-11 4:14 ` David Ahern
2015-02-06 15:10 ` Nicolas Dichtel
2015-02-06 20:50 ` Eric W. Biederman
2015-02-09 0:36 ` David Ahern
2015-02-09 11:30 ` Derek Fawcus
[not found] ` <871tlxtbhd.fsf_-_@x220.int.ebiederm.org>
2015-02-11 2:55 ` network namespace bloat Eric Dumazet
2015-02-11 3:18 ` Eric W. Biederman [this message]
2015-02-19 19:49 ` David Miller
2015-03-09 18:22 ` [PATCH net-next 0/6] tcp_metrics: Network namespace bloat reduction Eric W. Biederman
2015-03-09 18:27 ` [PATCH net-next 1/6] tcp_metrics: panic when tcp_metrics can not be allocated Eric W. Biederman
2015-03-09 18:50 ` Sergei Shtylyov
2015-03-11 19:22 ` Sergei Shtylyov
2015-03-09 18:27 ` [PATCH net-next 2/6] tcp_metrics: Mix the network namespace into the hash function Eric W. Biederman
2015-03-09 18:29 ` [PATCH net-next 3/6] tcp_metrics: Add a field tcpm_net and verify it matches on lookup Eric W. Biederman
2015-03-09 20:25 ` Julian Anastasov
2015-03-10 6:59 ` Eric W. Biederman
2015-03-10 8:23 ` Julian Anastasov
2015-03-11 0:58 ` Eric W. Biederman
2015-03-10 16:36 ` David Miller
2015-03-10 17:06 ` Eric W. Biederman
2015-03-10 17:29 ` David Miller
2015-03-10 17:56 ` Eric W. Biederman
2015-03-09 18:30 ` [PATCH net-next 4/6] tcp_metrics: Remove the unused return code from tcp_metrics_flush_all Eric W. Biederman
2015-03-09 18:30 ` [PATCH net-next 5/6] tcp_metrics: Rewrite tcp_metrics_flush_all Eric W. Biederman
2015-03-09 18:31 ` [PATCH net-next 6/6] tcp_metrics: Use a single hash table for all network namespaces Eric W. Biederman
2015-03-09 18:43 ` Eric Dumazet
2015-03-09 18:47 ` Eric Dumazet
2015-03-09 19:35 ` Eric W. Biederman
2015-03-09 20:21 ` Eric Dumazet
2015-03-09 20:09 ` [PATCH net-next 0/6] tcp_metrics: Network namespace bloat reduction David Miller
2015-03-09 20:21 ` Eric W. Biederman
2015-03-11 16:33 ` [PATCH net-next 0/8] tcp_metrics: Network namespace bloat reduction v2 Eric W. Biederman
2015-03-11 16:35 ` [PATCH net-next 1/8] net: Kill hold_net release_net Eric W. Biederman
2015-03-11 16:55 ` Eric Dumazet
2015-03-11 17:34 ` Eric W. Biederman
2015-03-11 17:07 ` Eric Dumazet
2015-03-11 17:08 ` Eric Dumazet
2015-03-11 17:10 ` Eric Dumazet
2015-03-11 17:36 ` Eric W. Biederman
2015-03-11 16:36 ` [PATCH net-next 2/8] net: Introduce possible_net_t Eric W. Biederman
2015-03-11 16:38 ` [PATCH net-next 3/8] tcp_metrics: panic when tcp_metrics_init fails Eric W. Biederman
2015-03-11 16:38 ` [PATCH net-next 4/8] tcp_metrics: Mix the network namespace into the hash function Eric W. Biederman
2015-03-11 16:40 ` [PATCH net-next 5/8] tcp_metrics: Add a field tcpm_net and verify it matches on lookup Eric W. Biederman
2015-03-11 16:41 ` [PATCH net-next 6/8] tcp_metrics: Remove the unused return code from tcp_metrics_flush_all Eric W. Biederman
2015-03-11 16:43 ` [PATCH net-next 7/8] tcp_metrics: Rewrite tcp_metrics_flush_all Eric W. Biederman
2015-03-11 16:43 ` [PATCH net-next 8/8] tcp_metrics: Use a single hash table for all network namespaces Eric W. Biederman
2015-03-13 5:04 ` [PATCH net-next 0/6] tcp_metrics: Network namespace bloat reduction v3 Eric W. Biederman
2015-03-13 5:04 ` [PATCH net-next 1/6] tcp_metrics: panic when tcp_metrics_init fails Eric W. Biederman
2015-03-13 5:05 ` [PATCH net-next 2/6] tcp_metrics: Mix the network namespace into the hash function Eric W. Biederman
2015-03-13 5:05 ` [PATCH net-next 3/6] tcp_metrics: Add a field tcpm_net and verify it matches on lookup Eric W. Biederman
2015-03-13 5:06 ` [PATCH net-next 4/6] tcp_metrics: Remove the unused return code from tcp_metrics_flush_all Eric W. Biederman
2015-03-13 5:07 ` [PATCH net-next 5/6] tcp_metrics: Rewrite tcp_metrics_flush_all Eric W. Biederman
2015-03-13 5:07 ` [PATCH net-next 6/6] tcp_metrics: Use a single hash table for all network namespaces Eric W. Biederman
2015-03-13 5:57 ` [PATCH net-next 0/6] tcp_metrics: Network namespace bloat reduction v3 David Miller
2015-02-11 17:09 ` network namespace bloat Nicolas Dichtel
2015-02-10 0:53 ` [RFC PATCH 00/29] net: VRF support Thomas Graf
2015-02-10 20:54 ` David Ahern
2016-05-25 16:04 ` Chenna
2016-05-25 19:04 ` David Ahern
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=87d25hrv9m.fsf@x220.int.ebiederm.org \
--to=ebiederm@xmission$(echo .)com \
--cc=davem@davemloft$(echo .)net \
--cc=ddutt@cumulusnetworks$(echo .)com \
--cc=dsahern@gmail$(echo .)com \
--cc=edumazet@google$(echo .)com \
--cc=hannes@stressinduktion$(echo .)org \
--cc=netdev@vger$(echo .)kernel.org \
--cc=nicolas.dichtel@6wind$(echo .)com \
--cc=roopa@cumulusnetworks$(echo .)com \
--cc=shmulik.ladkani@gmail$(echo .)com \
--cc=stephen@networkplumber$(echo .)org \
--cc=vipin@cumulusnetworks$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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