FS/XFS testing framework
 help / color / mirror / Atom feed
From: Dave Chinner <david@fromorbit.com>
To: fstests@vger.kernel.org
Cc: zlang@kernel.org
Subject: [PATCH 21/28] generic/531: limit max files per CPU
Date: Thu, 17 Apr 2025 13:01:02 +1000	[thread overview]
Message-ID: <20250417031208.1852171-22-david@fromorbit.com> (raw)
In-Reply-To: <20250417031208.1852171-1-david@fromorbit.com>

From: Dave Chinner <dchinner@redhat.com>

Currently g/531 runs t_open_files on every CPU, and with default
kernel settings that means 50,000 files per CPU are tested. On 64p
machines this means the test tries to create and unlink over 3
million files. This takes a long time:

Ten slowest tests - runtime in seconds:
generic/531 534
.....

Yet generic/531 is included in the 'quick' test group. It is
anything but "quick" on large CPU count systems.

Further, small filesystems  like are typically used for fstests do
not have the inherent concurrency to scale out this workload
effectively. Even using the mkfs.xfs concurrency options requires
using >250GB scratch devices on 64p machines because it won't make
AGs smaller than 4GB. Hence to get 64-way concurrency in the
filesystem, we need huge devices to be set up, and that's not really
practical for check-parallel.

Hence limit the total number of files this test will create
to a sane number, and distribute them over all the CPUs so that
the test runtime does not blow out on big systems. LOAD_FACTOR can
still be used to increase runtime of the test by increasing the
total number of files created.

Limiting the total number of files created brings g/531
system back into the "quick" test range on a 64p system:

generic/531        5s

Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
 tests/generic/531 | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tests/generic/531 b/tests/generic/531
index 07dffd9fd..3f691c0f8 100755
--- a/tests/generic/531
+++ b/tests/generic/531
@@ -29,13 +29,13 @@ _scratch_mount
 # Try to load up all the CPUs, two threads per CPU.
 nr_cpus=$(( $(getconf _NPROCESSORS_ONLN) * 2 ))
 
-# Set ULIMIT_NOFILE to min(file-max / $nr_cpus / 2, 50000 files per LOAD_FACTOR)
+# Set ULIMIT_NOFILE to min(file-max / 2, 100000) / $nr_cpus files per LOAD_FACTOR)
 # so that this test doesn't take forever or OOM the box
-max_files=$((50000 * LOAD_FACTOR))
-max_allowable_files=$(( $(cat /proc/sys/fs/file-max) / $nr_cpus / 2 ))
+max_files=$((100000 * LOAD_FACTOR))
+max_allowable_files=$(( $(cat /proc/sys/fs/file-max) / 2 ))
 test $max_allowable_files -gt 0 && test $max_files -gt $max_allowable_files && \
 	max_files=$max_allowable_files
-ulimit -n $max_files
+ulimit -n $((max_files / nr_cpus))
 
 # Open a lot of unlinked files
 echo create >> $seqres.full
-- 
2.45.2


  parent reply	other threads:[~2025-04-17  3:28 UTC|newest]

Thread overview: 80+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-17  3:00 [PATCH 00/28] check-parallel: Running tests without check Dave Chinner
2025-04-17  3:00 ` [PATCH 01/28] fstests: remove support for non-numeric test names Dave Chinner
2025-04-30  9:17   ` Nirjhar Roy (IBM)
2025-05-21  2:39     ` Dave Chinner
2025-05-26  5:14       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 02/28] _scratch_mkfs_sized: obey USE_EXTERNAL for XFS filesystems Dave Chinner
2025-05-05  6:14   ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 03/28] fstests: move test exit functions to common/exit Dave Chinner
2025-04-17  3:00 ` [PATCH 04/28] check-parallel: report how many tests were _notrun Dave Chinner
2025-05-05  9:58   ` Nirjhar Roy (IBM)
2025-05-21  2:53     ` Dave Chinner
2025-05-26  6:09       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 05/28] check: factor out test list building code Dave Chinner
2025-05-06 11:32   ` Nirjhar Roy (IBM)
2025-05-21  3:55     ` Dave Chinner
2025-05-26  6:48       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 06/28] check-parallel: use common group list parsing code Dave Chinner
2025-05-06 15:56   ` Nirjhar Roy (IBM)
2025-05-21  4:13     ` Dave Chinner
2025-05-26  6:58       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 07/28] check-parallel: adjust concurrency according to CPU count Dave Chinner
2025-05-07  6:45   ` Nirjhar Roy (IBM)
2025-05-21  4:32     ` Dave Chinner
2025-05-26  8:50       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 08/28] check-parallel: add logwrite device support Dave Chinner
2025-05-07  8:18   ` Nirjhar Roy (IBM)
2025-05-21 10:07     ` Dave Chinner
2025-05-26  8:59       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 09/28] check-parallel: allow FSTYP selection from the CLI Dave Chinner
2025-05-07  8:49   ` Nirjhar Roy (IBM)
2025-05-21 10:17     ` Dave Chinner
2025-05-26  9:00       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 10/28] check-parallel: use PID namespaces for runner process isolation Dave Chinner
2025-05-07  9:02   ` Nirjhar Roy (IBM)
2025-05-21 10:19     ` Dave Chinner
2025-05-26  9:04       ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 11/28] check-parallel: initial support for specifying device sizes Dave Chinner
2025-05-07 10:05   ` Nirjhar Roy (IBM)
2025-05-21 11:11     ` Dave Chinner
2025-04-17  3:00 ` [PATCH 12/28] config: move config section code to it's own file Dave Chinner
2025-05-09  6:09   ` Nirjhar Roy
2025-05-21 11:28     ` Dave Chinner
2025-04-17  3:00 ` [PATCH 13/28] check-parallel: introduce config file support Dave Chinner
2025-05-09 12:01   ` Nirjhar Roy
2025-05-21 12:23     ` Dave Chinner
2025-04-17  3:00 ` [PATCH 14/28] fstests: further separate sourcing common/rc and common/config from initialisation Dave Chinner
2025-05-10 14:08   ` Nirjhar Roy (IBM)
2025-04-17  3:00 ` [PATCH 15/28] check-parallel: de-batch test execution Dave Chinner
2025-05-09 13:16   ` Nirjhar Roy
2025-04-17  3:00 ` [PATCH 16/28] check-parallel: run sections directly Dave Chinner
2025-05-09 14:03   ` Nirjhar Roy
2025-04-17  3:00 ` [PATCH 17/28] check-parallel: rebuild test list when FSTYP changes Dave Chinner
2025-05-09 16:00   ` Nirjhar Roy
2025-04-17  3:00 ` [PATCH 18/28] check-parallel: create a "results-latest" symlink Dave Chinner
2025-05-10 13:12   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 19/28] check: factor test running Dave Chinner
2025-05-12 13:57   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 20/28] [RFC] check-parallel: run tests directly without using check Dave Chinner
2025-05-13 14:48   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` Dave Chinner [this message]
2025-05-10 13:15   ` [PATCH 21/28] generic/531: limit max files per CPU Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 22/28] fsync-tester.c: use syncfs() rather than sync() Dave Chinner
2025-04-30  9:08   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 23/28] open-by-handle.c: " Dave Chinner
2025-04-30  9:02   ` Nirjhar Roy (IBM)
2025-05-21  2:32     ` Dave Chinner
2025-05-26  5:11       ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 24/28] " Dave Chinner
2025-04-30  8:56   ` Nirjhar Roy (IBM)
2025-05-21  2:30     ` Dave Chinner
2025-05-26  4:56       ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 25/28] bulkstat_unlink_test_modified.c: remove unused test code Dave Chinner
2025-04-30  8:47   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 26/28] stale-handle.c: use syncfs() rather than sync() Dave Chinner
2025-04-30  8:34   ` Nirjhar Roy (IBM)
2025-05-21  2:24     ` Dave Chinner
2025-04-17  3:01 ` [PATCH 27/28] scaleread: remove dead test code Dave Chinner
2025-04-30  8:10   ` Nirjhar Roy (IBM)
2025-04-17  3:01 ` [PATCH 28/28] xfs/259: no need to call sync Dave Chinner
2025-04-30  7:56   ` Nirjhar Roy (IBM)

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=20250417031208.1852171-22-david@fromorbit.com \
    --to=david@fromorbit.com \
    --cc=fstests@vger.kernel.org \
    --cc=zlang@kernel.org \
    /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