diff mbox series

[v4,8/8] test-lib: add the '--stress' option to run a test repeatedly under load

Message ID 20190105010859.11031-9-szeder.dev@gmail.com (mailing list archive)
State New, archived
Headers show
Series [v4,1/8] test-lib: translate SIGTERM and SIGHUP to an exit | expand

Commit Message

SZEDER Gábor Jan. 5, 2019, 1:08 a.m. UTC
Unfortunately, we have a few flaky tests, whose failures tend to be
hard to reproduce.  We've found that the best we can do to reproduce
such a failure is to run the test script repeatedly while the machine
is under load, and wait in the hope that the load creates enough
variance in the timing of the test's commands that a failure is
evenually triggered.  I have a command to do that, and I noticed that
two other contributors have rolled their own scripts to do the same,
all choosing slightly different approaches.

To help reproduce failures in flaky tests, introduce the '--stress'
option to run a test script repeatedly in multiple parallel jobs until
one of them fails, thereby using the test script itself to increase
the load on the machine.

The number of parallel jobs is determined by, in order of precedence:
the number specified as '--stress=<N>', or the value of the
GIT_TEST_STRESS_LOAD environment variable, or twice the number of
available processors (as reported by the 'getconf' utility), or 8.

Make '--stress' imply '--verbose -x --immediate' to get the most
information about rare failures; there is really no point in spending
all the extra effort to reproduce such a failure, and then not know
which command failed and why.

To prevent the several parallel invocations of the same test from
interfering with each other:

  - Include the parallel job's number in the name of the trash
    directory and the various output files under 't/test-results/' as
    a '.stress-<Nr>' suffix.

  - Add the parallel job's number to the port number specified by the
    user or to the test number, so even tests involving daemons
    listening on a TCP socket can be stressed.

  - Redirect each parallel test run's verbose output to
    't/test-results/$TEST_NAME.stress-<nr>.out', because dumping the
    output of several parallel running tests to the terminal would
    create a big ugly mess.

For convenience, print the output of the failed test job at the end,
and rename its trash directory to end with the '.stress-failed'
suffix, so it's easy to find in a predictable path (OTOH, all absolute
paths recorded in the trash directory become invalid; we'll see
whether this causes any issues in practice).  If, in an unlikely case,
more than one jobs were to fail nearly at the same time, then print
the output of all failed jobs, and rename the trash directory of only
the last one (i.e. with the highest job number), as it is the trash
directory of the test whose output will be at the bottom of the user's

Based on Jeff King's 'stress' script.

Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
 t/README                |  19 ++++++-
 t/test-lib-functions.sh |   7 ++-
 t/test-lib.sh           | 109 +++++++++++++++++++++++++++++++++++++++-
 3 files changed, 130 insertions(+), 5 deletions(-)
diff mbox series


diff --git a/t/README b/t/README
index 28711cc508..11ce7675e3 100644
--- a/t/README
+++ b/t/README
@@ -186,6 +186,22 @@  appropriately before running "make".
 	this feature by setting the GIT_TEST_CHAIN_LINT environment
 	variable to "1" or "0", respectively.
+	Run the test script repeatedly in multiple parallel jobs until
+	one of them fails.  Useful for reproducing rare failures in
+	flaky tests.  The number of parallel jobs is, in order of
+	precedence: <N>, or the value of the GIT_TEST_STRESS_LOAD
+	environment variable, or twice the number of available
+	processors (as shown by the 'getconf' utility),	or 8.
+	Implies `--verbose -x --immediate` to get the most information
+	about the failure.  Note that the verbose output of each test
+	job is saved to 't/test-results/$TEST_NAME.stress-<nr>.out',
+	and only the output of the failed test job is shown on the
+	terminal.  The names of the trash directories get a
+	'.stress-<nr>' suffix, and the trash directory of the failed
+	test job is renamed to end with a '.stress-failed' suffix.
 You can also set the GIT_TEST_INSTALLED environment variable to
 the bindir of an existing git installation to test that installation.
 You still need to have built this git sandbox, from which various
@@ -425,7 +441,8 @@  This test harness library does the following things:
  - Creates an empty test directory with an empty .git/objects database
    and chdir(2) into it.  This directory is 't/trash
    directory.$test_name_without_dotsh', with t/ subject to change by
-   the --root option documented above.
+   the --root option documented above, and a '.stress-<N>' suffix
+   appended by the --stress option.
  - Defines standard test helper functions for your scripts to
    use.  These functions are designed to make all scripts behave
diff --git a/t/test-lib-functions.sh b/t/test-lib-functions.sh
index 4459bdda13..92cf8f812c 100644
--- a/t/test-lib-functions.sh
+++ b/t/test-lib-functions.sh
@@ -1288,8 +1288,6 @@  test_set_port () {
 			# root-only port, use a larger one instead.
 			port=$(($port + 10000))
-		eval $var=$port
 		error >&7 "invalid port number: $port"
@@ -1298,4 +1296,9 @@  test_set_port () {
 		# The user has specified the port.
+	# Make sure that parallel '--stress' test jobs get different
+	# ports.
+	port=$(($port + ${GIT_TEST_STRESS_JOB_NR:-0}))
+	eval $var=$port
diff --git a/t/test-lib.sh b/t/test-lib.sh
index 0e9ac9118d..a1abb1177a 100644
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -139,6 +139,19 @@  do
+	--stress)
+		stress=t ;;
+	--stress=*)
+		stress=${opt#--*=}
+		case "$stress" in
+		*[^0-9]*|0*|"")
+			echo "error: --stress=<N> requires the number of jobs to run" >&2
+			exit 1
+			;;
+		*)	# Good.
+			;;
+		esac
+		;;
 		echo "error: unknown test option '$opt'" >&2; exit 1 ;;
@@ -160,16 +173,108 @@  then
 	test -z "$verbose_log" && verbose=t
+if test -n "$stress"
+	verbose=t
+	trace=t
+	immediate=t
 TEST_NAME="$(basename "$0" .sh)"
-TRASH_DIRECTORY="trash directory.$TEST_NAME"
 test -n "$root" && TRASH_DIRECTORY="$root/$TRASH_DIRECTORY"
 /*) ;; # absolute path is good
+# If --stress was passed, run this test repeatedly in several parallel loops.
+if test "$GIT_TEST_STRESS_STARTED" = "done"
+	: # Don't stress test again.
+elif test -n "$stress"
+	if test "$stress" != t
+	then
+		job_count=$stress
+	elif test -n "$GIT_TEST_STRESS_LOAD"
+	then
+		job_count="$GIT_TEST_STRESS_LOAD"
+	elif job_count=$(getconf _NPROCESSORS_ONLN 2>/dev/null) &&
+	     test -n "$job_count"
+	then
+		job_count=$((2 * $job_count))
+	else
+		job_count=8
+	fi
+	mkdir -p "$TEST_RESULTS_DIR"
+	stressfail="$TEST_RESULTS_BASE.stress-failed"
+	rm -f "$stressfail"
+	stress_exit=0
+	trap '
+		kill $job_pids 2>/dev/null
+		wait
+		stress_exit=1
+	job_pids=
+	job_nr=0
+	while test $job_nr -lt "$job_count"
+	do
+		(
+			trap '
+				kill $test_pid 2>/dev/null
+				wait
+				exit 1
+			' TERM INT
+			cnt=0
+			while ! test -e "$stressfail"
+			do
+				$TEST_SHELL_PATH "$0" "$@" >"$TEST_RESULTS_BASE.stress-$job_nr.out" 2>&1 &
+				test_pid=$!
+				if wait $test_pid
+				then
+					printf "OK   %2d.%d\n" $GIT_TEST_STRESS_JOB_NR $cnt
+				else
+					echo $GIT_TEST_STRESS_JOB_NR >>"$stressfail"
+					printf "FAIL %2d.%d\n" $GIT_TEST_STRESS_JOB_NR $cnt
+				fi
+				cnt=$(($cnt + 1))
+			done
+		) &
+		job_pids="$job_pids $!"
+		job_nr=$(($job_nr + 1))
+	done
+	wait
+	if test -f "$stressfail"
+	then
+		echo "Log(s) of failed test run(s):"
+		for failed_job_nr in $(sort -n "$stressfail")
+		do
+			echo "Contents of '$TEST_RESULTS_BASE.stress-$failed_job_nr.out':"
+			cat "$TEST_RESULTS_BASE.stress-$failed_job_nr.out"
+		done
+		rm -rf "$TRASH_DIRECTORY.stress-failed"
+		# Move the last one.
+		mv "$TRASH_DIRECTORY.stress-$failed_job_nr" "$TRASH_DIRECTORY.stress-failed"
+	fi
+	exit $stress_exit
 # if --tee was passed, write the output not only to the terminal, but
 # additionally to the file test-results/$BASENAME.out, too.
 if test "$GIT_TEST_TEE_STARTED" = "done"