From 40abc63c4043a6b51a9ab656d27d88503cc775d8 Mon Sep 17 00:00:00 2001
From: Brian Behlendorf <behlendorf1@llnl.gov>
Date: Sun, 18 Apr 2021 21:58:36 -0700
Subject: [PATCH] ZTS: Improve redundancy test scripts

- Add additional logging to provide more information about why the
  test failed.  This including logging more of the individual commands
  and the contents and differences of the record files on failure.

- Updated get_vdevs() to properly exclude all top-level vdevs
  including raidz3 and draid[1-3].

- Replaced gnudd with dd.  This is the only remaining place in the
  test suite gnudd is used and it shouldn't be needed.

- The refill_test_env function expects the pool as the first argument
  but never sets the pool variable.

- Only fill the test pools to 50% of capacity instead of 75% to help
  speed up the tests.

- Fix replace_missing_devs() calculation, MINDEVSIZE should be
  MINVDEVSIZE.

- Fix damage_devs() so it overwrites almost all of the device so
  we're guaranteed to damage filesystem blocks.

- redundancy_stripe.ksh should not use log_mustnot to check if the
  pool is healthy since the return value may be misinterpreted.
  Just perform a normal conditional check and log the failure.

Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #11906
---
 .../functional/redundancy/redundancy.kshlib   | 24 ++++++++++++-------
 .../redundancy/redundancy_draid3.ksh          |  2 +-
 .../redundancy/redundancy_stripe.ksh          |  4 +++-
 3 files changed, 19 insertions(+), 11 deletions(-)

diff --git a/tests/zfs-tests/tests/functional/redundancy/redundancy.kshlib b/tests/zfs-tests/tests/functional/redundancy/redundancy.kshlib
index 26ded8720..baee8269b 100644
--- a/tests/zfs-tests/tests/functional/redundancy/redundancy.kshlib
+++ b/tests/zfs-tests/tests/functional/redundancy/redundancy.kshlib
@@ -146,7 +146,7 @@ function setup_test_env
 	typeset -i i=0
 	typeset file=$TESTDIR/file
 	typeset -i limit
-	(( limit = $(get_prop available $pool) / 4 ))
+	(( limit = $(get_prop available $pool) / 2 ))
 
 	while true ; do
 		[[ $(get_prop available $pool) -lt $limit ]] && break
@@ -162,6 +162,7 @@ function setup_test_env
 function refill_test_env
 {
 	log_note "Re-filling the filesystem ..."
+	typeset pool=$1
 	typeset -i ret=0
 	typeset -i i=0
 	typeset mntpnt
@@ -217,8 +218,13 @@ function is_data_valid
 {
 	typeset pool=$1
 
+	log_must zpool scrub -w $pool
+
 	record_data $pool $PST_RECORD_FILE
 	if ! diff $PRE_RECORD_FILE $PST_RECORD_FILE > /dev/null 2>&1; then
+		log_must cat $PRE_RECORD_FILE
+		log_must cat $PST_RECORD_FILE
+		diff -u $PRE_RECORD_FILE $PST_RECORD_FILE
 		return 1
 	fi
 
@@ -237,7 +243,7 @@ function get_vdevs #pool cnt
 	typeset -i cnt=$2
 
 	typeset all_devs=$(zpool iostat -v $pool | awk '{print $1}'| \
-		egrep -v "^pool$|^capacity$|^mirror$|^raidz1$|^raidz2$|---" | \
+		egrep -v "^pool$|^capacity$|^mirror$|^raidz1$|^raidz2$|^raidz3$|^draid1.*|^draid2.*|^draid3.*|---" | \
 		egrep -v "/old$|^$pool$")
 	typeset -i i=0
 	typeset vdevs
@@ -265,9 +271,9 @@ function replace_missing_devs
 
 	typeset vdev
 	for vdev in $@; do
-		log_must gnudd if=/dev/zero of=$vdev \
-		    bs=1024k count=$(($MINDEVSIZE / (1024 * 1024))) \
-		    oflag=fdatasync
+		log_must dd if=/dev/zero of=$vdev \
+		    bs=1024k count=$((MINVDEVSIZE / (1024 * 1024))) \
+		    conv=fdatasync
 		log_must zpool replace -wf $pool $vdev $vdev
 	done
 }
@@ -286,19 +292,19 @@ function damage_devs
 	typeset -i cnt=$2
 	typeset label="$3"
 	typeset vdevs
-	typeset -i bs_count=$((64 * 1024))
+	typeset -i bs_count=$(((MINVDEVSIZE / 1024) - 4096))
 
 	vdevs=$(get_vdevs $pool $cnt)
 	typeset dev
 	if [[ -n $label ]]; then
 		for dev in $vdevs; do
-			dd if=/dev/zero of=$dev seek=512 bs=1024 \
+			log_must dd if=/dev/zero of=$dev seek=512 bs=1024 \
 			    count=$bs_count conv=notrunc >/dev/null 2>&1
 		done
 	else
 		for dev in $vdevs; do
-			dd if=/dev/zero of=$dev bs=1024 count=$bs_count \
-			    conv=notrunc >/dev/null 2>&1
+			log_must dd if=/dev/zero of=$dev bs=1024 \
+			    count=$bs_count conv=notrunc >/dev/null 2>&1
 		done
 	fi
 
diff --git a/tests/zfs-tests/tests/functional/redundancy/redundancy_draid3.ksh b/tests/zfs-tests/tests/functional/redundancy/redundancy_draid3.ksh
index bddd150d0..d4c823ed9 100755
--- a/tests/zfs-tests/tests/functional/redundancy/redundancy_draid3.ksh
+++ b/tests/zfs-tests/tests/functional/redundancy/redundancy_draid3.ksh
@@ -42,7 +42,7 @@
 #	2. Create draid3 pool based on the virtual disk files.
 #	3. Fill the filesystem with directories and files.
 #	4. Record all the files and directories checksum information.
-#	5. Damaged at most two of the virtual disk files.
+#	5. Damaged at most three of the virtual disk files.
 #	6. Verify the data is correct to prove draid3 can withstand 3 devices
 #	   are failing.
 #
diff --git a/tests/zfs-tests/tests/functional/redundancy/redundancy_stripe.ksh b/tests/zfs-tests/tests/functional/redundancy/redundancy_stripe.ksh
index 7ee51051e..b2c4a85fe 100755
--- a/tests/zfs-tests/tests/functional/redundancy/redundancy_stripe.ksh
+++ b/tests/zfs-tests/tests/functional/redundancy/redundancy_stripe.ksh
@@ -57,6 +57,8 @@ setup_test_env $TESTPOOL "" $cnt
 damage_devs $TESTPOOL 1 "keep_label"
 log_must zpool scrub -w $TESTPOOL
 
-log_mustnot is_healthy $TESTPOOL
+if is_healthy $TESTPOOL ; then
+	log_fail "$pool should not be healthy."
+fi
 
 log_pass "Striped pool has no data redundancy as expected."