Date: Fri, 18 Jun 2021 14:43:48 GMT From: Warner Losh <imp@FreeBSD.org> To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-main@FreeBSD.org Subject: git: 0f882bb14a0e - main - NOTES: Update the HZ entry with latest advice. Message-ID: <202106181443.15IEhm3h010711@gitrepo.freebsd.org>
next in thread | raw e-mail | index | archive | help
The branch main has been updated by imp: URL: https://cgit.FreeBSD.org/src/commit/?id=0f882bb14a0e772bf901b1a3e40d028a4165fbd9 commit 0f882bb14a0e772bf901b1a3e40d028a4165fbd9 Author: Warner Losh <imp@FreeBSD.org> AuthorDate: 2021-06-18 14:39:18 +0000 Commit: Warner Losh <imp@FreeBSD.org> CommitDate: 2021-06-18 14:41:56 +0000 NOTES: Update the HZ entry with latest advice. Sponsored by: Netflix Differential Revision: https://reviews.freebsd.org/D30804 --- sys/conf/NOTES | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/sys/conf/NOTES b/sys/conf/NOTES index 40063c09382f..89dd572986bc 100644 --- a/sys/conf/NOTES +++ b/sys/conf/NOTES @@ -1047,8 +1047,7 @@ options TCP_SIGNATURE #include support for RFC 2385 # DUMMYNET enables the "dummynet" bandwidth limiter. You need IPFIREWALL # as well. See dummynet(4) and ipfw(8) for more info. When you run -# DUMMYNET it is advisable to also have at least "options HZ=1000" to achieve -# a smooth scheduling of the traffic. +# DUMMYNET, HZ/kern.hz should be at least 1000 for adequate response. options DUMMYNET # The DEBUGNET option enables a basic debug/panic-time networking API. It @@ -1237,14 +1236,11 @@ options CAPABILITY_MODE # sandboxes with no global namespace access ##################################################################### # CLOCK OPTIONS -# The granularity of operation is controlled by the kernel option HZ whose -# default value (1000 on most architectures) means a granularity of 1ms -# (1s/HZ). Historically, the default was 100, but finer granularity is -# required for DUMMYNET and other systems on modern hardware. There are -# reasonable arguments that HZ should, in fact, be 100 still; consider, -# that reducing the granularity too much might cause excessive overhead in -# clock interrupt processing, potentially causing ticks to be missed and thus -# actually reducing the accuracy of operation. +# The granularity of operation is controlled by the kernel option HZ (default +# 1000 or 1ms). Virtual machines guests default to an HZ value of 100. Lower +# values produce less overhead at the expense of reduce accuracy in delivery of +# kernel timeouts. With the adaptive tick code, lower values produce less +# benefit than in the past. options HZ=100
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?202106181443.15IEhm3h010711>