Thread (9 messages) flat view 9 messages, 2 authors, 1d ago
WARM1d

[PATCH net-next RFC 0/6] netconsole: Support messages ratelimit-ing

From: Breno Leitao <leitao@debian.org>
Date: 2026-08-18 10:29:47
Also in: linux-doc, linux-kselftest, lkml

netconsole hands every console message to every enabled target, with no
bound on the rate.

At Meta, a few hosts caused some DoS and netconsd was OOM killed in
some regions, fed by 3 billion RCU messages from thousand hosts and by
a swap fault that retries forever, which logged 6 million copies of one
line on a single host in a few hours.

Both floods are being fixed where they are printed, by Paula and me for RCU
Tasks [1][2][3] and by me for swap faul fix [4], but that only ever
covers the flood already found, not new issues that might be happening.

Use the ratelimit in the kernel to ratelimit messages from netconsole.

Give each target a token bucket, consulted once per message so that a
message split into several ncfrag packets is sent whole or dropped
whole.

It starts with a zero interval, which struct ratelimit_state treats as
unlimited, so nothing changes until ratelimit_interval_ms and
ratelimit_burst are written.

What the bucket discards is reported on the wire, so the receiver sees
"netconsole: 45 messages dropped by rate limit" rather than an
unexplained gap. This is what is captured on netconsd, when this patch
is applied:

	7.2.0-rc7-01460-g75848acaf136,13,628,252392647,-;netconsole selftest: netcons_AGujw 1
	7.2.0-rc7-01460-g75848acaf136,13,629,252392782,-;netconsole selftest: netcons_AGujw 2
	....
	7.2.0-rc7-01460-g75848acaf136,4,0,254742176,-;netconsole: 45 messages dropped by rate limit
	7.2.0-rc7-01460-g75848acaf136,13,678,254742125,-;netconsole selftest: netcons_AGujw 1

Crash output is exempt. The bucket is skipped while oops_in_progress is
set, so a limit configured for steady-state logging cannot cost a target
part of an oops or a panic.

[1] https://lore.kernel.org/all/anw8Qw8lfeIykski@gmail.com/ (local)
[2] https://lore.kernel.org/all/20260810-rcu_task_shrink_lruvec-v1-1-4d9f7d5251cb@debian.org/ (local)
[3] https://lore.kernel.org/all/20260810-rcu_task_shrink_lruvec-v1-1-4d9f7d5251cb@debian.org/ (local)
[4] https://lore.kernel.org/all/20260813-swap-v2-0-4a625ccabdae@debian.org/ (local)

Signed-off-by: Breno Leitao <leitao@debian.org>
---
Breno Leitao (6):
      netconsole: add a per-target message rate limit
      netconsole: allow configuring the rate limit interval through configfs
      netconsole: allow configuring the rate limit burst through configfs
      netconsole: tell the target when the rate limit drops messages
      docs: netconsole: document rate limit feature
      selftests: netcons: test the per-target rate limit

 Documentation/networking/netconsole.rst            |  43 ++++++
 drivers/net/netconsole.c                           | 145 +++++++++++++++++++
 .../selftests/drivers/net/netconsole/Makefile      |   1 +
 .../drivers/net/netconsole/netcons_ratelimit.sh    | 160 +++++++++++++++++++++
 4 files changed, 349 insertions(+)
---
base-commit: e6a5d573d24cd375e09d24f136523cb3cc85c9d3
change-id: 20260817-netcons_ratelimit-629b04a73c57

Best regards,
--  
Breno Leitao [off-list ref]
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help