Systemd: Remove 'use-bfq-scheduler.patch.'
Let's revert to 'mq_deadline' as the default I/O scheduler. BFQ is causing regressions in MySQL + vSAN storage. Change-Id: I58a507e2f26f8ebccb071753f4aae0d0618fcf8c Signed-off-by: Guruswamy Basavaiah <guruswamy.basavaiah@broadcom.com> Reviewed-on: http://photon-jenkins.eng.vmware.com:8082/c/photon/+/23353 Reviewed-by: Shreenidhi Shedi <shreenidhi.shedi@broadcom.com> Tested-by: gerrit-photon <photon-checkins@vmware.com>
This commit is contained in:
parent
f4190cc956
commit
990b3fa0d6
|
@ -3,7 +3,7 @@
|
|||
Name: systemd
|
||||
URL: http://www.freedesktop.org/wiki/Software/systemd
|
||||
Version: 255.2
|
||||
Release: 1%{?dist}
|
||||
Release: 2%{?dist}
|
||||
License: LGPLv2+ and GPLv2+ and MIT
|
||||
Summary: System and Service Manager
|
||||
Group: System Environment/Security
|
||||
|
@ -28,7 +28,6 @@ Source14: sysusers.generate-pre.sh
|
|||
|
||||
Patch0: enoX-uses-instance-number-for-vmware-hv.patch
|
||||
Patch1: fetch-dns-servers-from-environment.patch
|
||||
Patch2: use-bfq-scheduler.patch
|
||||
|
||||
Requires: Linux-PAM
|
||||
Requires: bzip2
|
||||
|
@ -754,6 +753,8 @@ fi
|
|||
%files lang -f ../%{name}.lang
|
||||
|
||||
%changelog
|
||||
* Tue Feb 27 2024 Susant Sahani <guruswamy.basavaiah@broadcom.com> 255.2-2
|
||||
- use-bfq-scheduler.patch
|
||||
* Thu Jan 04 2024 Susant Sahani <susant.sahani@broadcom.com> 255.2-1
|
||||
- Version bump.
|
||||
* Fri Nov 24 2023 Shreenidhi Shedi <sshedi@vmware.com> 254.1-8
|
||||
|
|
|
@ -1,41 +0,0 @@
|
|||
From 49ac7437fc8e1d002c5588015babc1abeaa5b990 Mon Sep 17 00:00:00 2001
|
||||
From: =?UTF-8?q?Zbigniew=20J=C4=99drzejewski-Szmek?= <zbyszek@in.waw.pl>
|
||||
Date: Tue, 2 Jan 2024 15:05:00 +0000
|
||||
Subject: [PATCH] udev: use bfq as the default scheduler
|
||||
|
||||
As requested in https://bugzilla.redhat.com/show_bug.cgi?id=1738828.
|
||||
Test results are that bfq seems to behave better and more consistently on
|
||||
typical hardware. The kernel does not have a configuration option to set
|
||||
the default scheduler, and it currently needs to be set by userspace.
|
||||
|
||||
See the bug for more discussion and links.
|
||||
---
|
||||
rules.d/60-block-scheduler.rules | 3 +++
|
||||
rules.d/meson.build | 1 +
|
||||
2 files changed, 4 insertions(+)
|
||||
create mode 100644 rules.d/60-block-scheduler.rules
|
||||
|
||||
diff --git a/rules.d/60-block-scheduler.rules b/rules.d/60-block-scheduler.rules
|
||||
new file mode 100644
|
||||
index 0000000..30a4f4c
|
||||
--- /dev/null
|
||||
+++ b/rules.d/60-block-scheduler.rules
|
||||
@@ -0,0 +1,3 @@
|
||||
+ACTION=="add", SUBSYSTEM=="block", ENV{DEVTYPE}=="disk", \
|
||||
+ KERNEL=="mmcblk*[0-9]|msblk*[0-9]|mspblk*[0-9]|sd*[!0-9]|sr*", \
|
||||
+ ATTR{queue/scheduler}="bfq"
|
||||
diff --git a/rules.d/meson.build b/rules.d/meson.build
|
||||
index 3040fae..852a892 100644
|
||||
--- a/rules.d/meson.build
|
||||
+++ b/rules.d/meson.build
|
||||
@@ -8,6 +8,7 @@ rules = [
|
||||
[files('60-autosuspend.rules',
|
||||
'60-block.rules',
|
||||
'60-cdrom_id.rules',
|
||||
+ '60-block-scheduler.rules',
|
||||
'60-dmi-id.rules',
|
||||
'60-drm.rules',
|
||||
'60-evdev.rules',
|
||||
--
|
||||
2.39.0
|
||||
|
Loading…
Reference in New Issue