From: Ming Lei <ming.lei@redhat.com>
To: Omar Sandoval <osandov@fb.com>
Cc: Jens Axboe <axboe@kernel.dk>,
linux-block@vger.kernel.org, Ming Lei <ming.lei@redhat.com>
Subject: [PATCH] block/023: performance test on queue creation & cleanup
Date: Wed, 4 Jul 2018 13:29:56 +0800 [thread overview]
Message-ID: <20180704052956.5110-1-ming.lei@redhat.com> (raw)
SCSI may have lots of channels, targets or LUNs, so it may
take long time for creating and cleaning up queues.
So introduce block/023 and uses null_blk to run this test
on both blk-mq and legacy mode, then compare both and check
the difference.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
---
tests/block/023 | 101 ++++++++++++++++++++++++++++++++++++++++++++++++++++
tests/block/023.out | 2 ++
2 files changed, 103 insertions(+)
create mode 100755 tests/block/023
create mode 100755 tests/block/023.out
diff --git a/tests/block/023 b/tests/block/023
new file mode 100755
index 000000000000..5a8f323bd42a
--- /dev/null
+++ b/tests/block/023
@@ -0,0 +1,101 @@
+#!/bin/bash
+#
+# Performance regression test on blk-mq queue creation & cleanup. Measure
+# the time for creating & cleaning up null_blk queues in both legacy and
+# blk-mq mode, then compare both & check if there is performance regression.
+#
+# Copyright (C) 2018 Ming Lei
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see <http://www.gnu.org/licenses/>.
+
+. tests/block/rc
+
+DESCRIPTION="performance regression test on null_blk queue creation & cleanup"
+QUICK=1
+
+requires() {
+ _have_module null_blk
+}
+
+remove_null_blk_mod() {
+ delay=$1
+ while true; do
+ modprobe -r null_blk > /dev/null 2>&1
+ [ $? -eq 0 ] && break
+ sleep $delay
+ done
+}
+
+test_null_blk_queue_perf() {
+ QUEUE_MODE=$1
+
+ remove_null_blk_mod 1
+
+ start_time=`date +'%s'`
+ modprobe null_blk nr_devices=4096 queue_mode=$QUEUE_MODE
+ end_time=`date +'%s'`
+ create_time=`expr $end_time - $start_time + 1`
+
+ start_time=`date +'%s'`
+ remove_null_blk_mod 0.1
+ end_time=`date +'%s'`
+ cleanup_time=`expr $end_time - $start_time + 1`
+
+ echo $create_time $cleanup_time
+}
+
+check_diff() {
+ src=$1
+ dst=$2
+ msg=$3
+
+ diff=`expr $dst - $src`
+ [ $diff -le 10 ] && return
+
+ result=`echo "scale=0;($diff - $src * 0.4) / 1" | bc`
+ [ $result -ge 0 ] && echo $msg
+}
+
+check_limit() {
+ src=$1
+ dst=$2
+ msg=$3
+
+ result=`echo "scale=0;$dst - $src * 20" | bc`
+ [ $result -ge 0 ] && echo $msg
+}
+
+test() {
+ echo "Running ${TEST_NAME}"
+
+ rq_time=($(test_null_blk_queue_perf 1))
+ mq_time=($(test_null_blk_queue_perf 2))
+
+ echo ${rq_time[0]} ${mq_time[0]} >> "$FULL"
+ echo ${rq_time[1]} ${mq_time[1]} >> "$FULL"
+
+ msg="blk_mq takes too long(${mq_time[0]}s) to create queues compared with block rq(${rq_time[0]}s)"
+ check_diff ${rq_time[0]} ${mq_time[0]} "$msg"
+
+ msg="blk_mq takes too long(${mq_time[1]}s) to cleanup queues compared with block rq(${rq_time[1]}s)"
+ check_diff ${rq_time[1]} ${mq_time[1]} "$msg"
+
+ msg="block rq takes too long(${rq_time[1]}s) to cleanup queues compared with creating queues(${rq_time[0]}s)"
+ check_limit ${rq_time[0]} ${rq_time[1]} "$msg"
+
+ msg="blk_mq takes too long(${mq_time[1]}s) to cleanup queues compared with creating queues(${mq_time[0]}s)"
+ check_limit ${mq_time[0]} ${mq_time[1]} "$msg"
+
+ echo "Test complete"
+}
diff --git a/tests/block/023.out b/tests/block/023.out
new file mode 100755
index 000000000000..4a11db1f10fe
--- /dev/null
+++ b/tests/block/023.out
@@ -0,0 +1,2 @@
+Running block/023
+Test complete
--
2.9.5
next reply other threads:[~2018-07-04 5:29 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-04 5:29 Ming Lei [this message]
2018-07-25 21:33 ` [PATCH] block/023: performance test on queue creation & cleanup Omar Sandoval
2018-07-29 22:43 ` Ming Lei
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=20180704052956.5110-1-ming.lei@redhat.com \
--to=ming.lei@redhat.com \
--cc=axboe@kernel.dk \
--cc=linux-block@vger.kernel.org \
--cc=osandov@fb.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