From: Aaron Conole <aconole@redhat.com>
To: Adrian Moreno <amorenoz@redhat.com>
Cc: netdev@vger.kernel.org, echaudro@redhat.com, horms@kernel.org,
i.maximets@ovn.org, dev@openvswitch.org,
Ido Schimmel <idosch@nvidia.com>,
Yotam Gigi <yotam.gi@gmail.com>,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>,
Paolo Abeni <pabeni@redhat.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH net-next v7 04/10] net: psample: allow using rate as probability
Date: Mon, 01 Jul 2024 14:10:34 -0400 [thread overview]
Message-ID: <f7tplrxvv8l.fsf@redhat.com> (raw)
In-Reply-To: <20240630195740.1469727-5-amorenoz@redhat.com> (Adrian Moreno's message of "Sun, 30 Jun 2024 21:57:25 +0200")
Adrian Moreno <amorenoz@redhat.com> writes:
> Although not explicitly documented in the psample module itself, the
> definition of PSAMPLE_ATTR_SAMPLE_RATE seems inherited from act_sample.
>
> Quoting tc-sample(8):
> "RATE of 100 will lead to an average of one sampled packet out of every
> 100 observed."
>
> With this semantics, the rates that we can express with an unsigned
> 32-bits number are very unevenly distributed and concentrated towards
> "sampling few packets".
> For example, we can express a probability of 2.32E-8% but we
> cannot express anything between 100% and 50%.
>
> For sampling applications that are capable of sampling a decent
> amount of packets, this sampling rate semantics is not very useful.
>
> Add a new flag to the uAPI that indicates that the sampling rate is
> expressed in scaled probability, this is:
> - 0 is 0% probability, no packets get sampled.
> - U32_MAX is 100% probability, all packets get sampled.
>
> Acked-by: Eelco Chaudron <echaudro@redhat.com>
> Reviewed-by: Ido Schimmel <idosch@nvidia.com>
> Signed-off-by: Adrian Moreno <amorenoz@redhat.com>
> ---
Reviewed-by: Aaron Conole <aconole@redhat.com>
> include/net/psample.h | 3 ++-
> include/uapi/linux/psample.h | 10 +++++++++-
> net/psample/psample.c | 3 +++
> 3 files changed, 14 insertions(+), 2 deletions(-)
>
> diff --git a/include/net/psample.h b/include/net/psample.h
> index 2ac71260a546..c52e9ebd88dd 100644
> --- a/include/net/psample.h
> +++ b/include/net/psample.h
> @@ -24,7 +24,8 @@ struct psample_metadata {
> u8 out_tc_valid:1,
> out_tc_occ_valid:1,
> latency_valid:1,
> - unused:5;
> + rate_as_probability:1,
> + unused:4;
> const u8 *user_cookie;
> u32 user_cookie_len;
> };
> diff --git a/include/uapi/linux/psample.h b/include/uapi/linux/psample.h
> index e80637e1d97b..b765f0e81f20 100644
> --- a/include/uapi/linux/psample.h
> +++ b/include/uapi/linux/psample.h
> @@ -8,7 +8,11 @@ enum {
> PSAMPLE_ATTR_ORIGSIZE,
> PSAMPLE_ATTR_SAMPLE_GROUP,
> PSAMPLE_ATTR_GROUP_SEQ,
> - PSAMPLE_ATTR_SAMPLE_RATE,
> + PSAMPLE_ATTR_SAMPLE_RATE, /* u32, ratio between observed and
> + * sampled packets or scaled probability
> + * if PSAMPLE_ATTR_SAMPLE_PROBABILITY
> + * is set.
> + */
> PSAMPLE_ATTR_DATA,
> PSAMPLE_ATTR_GROUP_REFCOUNT,
> PSAMPLE_ATTR_TUNNEL,
> @@ -20,6 +24,10 @@ enum {
> PSAMPLE_ATTR_TIMESTAMP, /* u64, nanoseconds */
> PSAMPLE_ATTR_PROTO, /* u16 */
> PSAMPLE_ATTR_USER_COOKIE, /* binary, user provided data */
> + PSAMPLE_ATTR_SAMPLE_PROBABILITY,/* no argument, interpret rate in
> + * PSAMPLE_ATTR_SAMPLE_RATE as a
> + * probability scaled 0 - U32_MAX.
> + */
>
> __PSAMPLE_ATTR_MAX
> };
> diff --git a/net/psample/psample.c b/net/psample/psample.c
> index 1c76f3e48dcd..f48b5b9cd409 100644
> --- a/net/psample/psample.c
> +++ b/net/psample/psample.c
> @@ -497,6 +497,9 @@ void psample_sample_packet(struct psample_group *group, struct sk_buff *skb,
> md->user_cookie))
> goto error;
>
> + if (md->rate_as_probability)
> + nla_put_flag(skb, PSAMPLE_ATTR_SAMPLE_PROBABILITY);
> +
> genlmsg_end(nl_skb, data);
> genlmsg_multicast_netns(&psample_nl_family, group->net, nl_skb, 0,
> PSAMPLE_NL_MCGRP_SAMPLE, GFP_ATOMIC);
next prev parent reply other threads:[~2024-07-01 18:10 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-30 19:57 [PATCH net-next v7 00/10] net: openvswitch: Add sample multicasting Adrian Moreno
2024-06-30 19:57 ` [PATCH net-next v7 01/10] net: psample: add user cookie Adrian Moreno
2024-07-01 11:20 ` Michal Kubiak
2024-07-01 18:05 ` Aaron Conole
2024-06-30 19:57 ` [PATCH net-next v7 02/10] net: sched: act_sample: add action cookie to sample Adrian Moreno
2024-07-01 18:06 ` Aaron Conole
2024-06-30 19:57 ` [PATCH net-next v7 03/10] net: psample: skip packet copy if no listeners Adrian Moreno
2024-07-01 18:07 ` Aaron Conole
2024-06-30 19:57 ` [PATCH net-next v7 04/10] net: psample: allow using rate as probability Adrian Moreno
2024-07-01 18:10 ` Aaron Conole [this message]
2024-06-30 19:57 ` [PATCH net-next v7 05/10] net: openvswitch: add psample action Adrian Moreno
2024-07-01 11:40 ` Michal Kubiak
2024-07-01 12:56 ` Adrián Moreno
2024-07-02 11:10 ` Michal Kubiak
2024-07-01 18:13 ` Ilya Maximets
2024-07-01 18:23 ` Aaron Conole
2024-07-02 7:05 ` Adrián Moreno
2024-07-02 7:29 ` Adrián Moreno
2024-07-02 9:37 ` Simon Horman
2024-07-02 9:52 ` Adrián Moreno
2024-07-02 9:53 ` Ilya Maximets
2024-07-02 12:33 ` [ovs-dev] " Simon Horman
2024-07-02 18:37 ` Adrián Moreno
2024-07-02 18:06 ` Jakub Kicinski
2024-07-02 18:16 ` Ilya Maximets
2024-07-02 18:24 ` Jakub Kicinski
2024-07-02 18:52 ` Ilya Maximets
2024-06-30 19:57 ` [PATCH net-next v7 06/10] net: openvswitch: store sampling probability in cb Adrian Moreno
2024-07-01 18:24 ` Aaron Conole
2024-06-30 19:57 ` [PATCH net-next v7 07/10] selftests: openvswitch: add psample action Adrian Moreno
2024-07-01 18:40 ` Aaron Conole
2024-06-30 19:57 ` [PATCH net-next v7 08/10] selftests: openvswitch: add userspace parsing Adrian Moreno
2024-06-30 19:57 ` [PATCH net-next v7 09/10] selftests: openvswitch: parse trunc action Adrian Moreno
2024-06-30 19:57 ` [PATCH net-next v7 10/10] selftests: openvswitch: add psample test Adrian Moreno
2024-07-01 18:34 ` Ilya Maximets
2024-07-01 18:38 ` Aaron Conole
2024-07-02 7:16 ` Adrián Moreno
2024-07-02 11:45 ` Aaron Conole
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=f7tplrxvv8l.fsf@redhat.com \
--to=aconole@redhat.com \
--cc=amorenoz@redhat.com \
--cc=davem@davemloft.net \
--cc=dev@openvswitch.org \
--cc=echaudro@redhat.com \
--cc=edumazet@google.com \
--cc=horms@kernel.org \
--cc=i.maximets@ovn.org \
--cc=idosch@nvidia.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=yotam.gi@gmail.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).