tests: improve and fix some test scripts

Change-Id: Iceefe22af754096c599dc570d4894d14fce4deae
Updates: bz#1193929
Signed-off-by: Xavier Hernandez <xhernandez@redhat.com>

9 files changed, 76 insertions, 27 deletions

diff --git a/tests/bugs/distribute/bug-1161311.t b/tests/bugs/distribute/bug-1161311.t
index 9ae4fac342f..62796068928 100755
--- a/tests/bugs/distribute/bug-1161311.t
+++ b/tests/bugs/distribute/bug-1161311.t
@@ -76,14 +76,12 @@ TEST glusterfs -s $H0 --volfile-id $V0 $M0;
 TEST mkdir $M0/dir1
 TEST mkdir -p $M0/dir2/dir3
 
-# Create a large file (6.4 GB), so that rebalance takes time
-# Reading from /dev/urandom is slow, so we'll cat it together
-dd if=/dev/urandom of=/tmp/FILE2 bs=64k count=10240
-for i in {1..10}; do
-  cat /tmp/FILE2 >> $M0/dir1/FILE2
-done
-
-#dd if=/dev/urandom of=$M0/dir1/FILE2 bs=64k count=10240
+# Create a large file (8 GB), so that rebalance takes time
+# Since we really don't care about the contents of the file, we use fallocate
+# to generate the file much faster. We could also use truncate, which is even
+# faster, but rebalance could take advantage of an sparse file and migrate it
+# in an optimized way, but we don't want a fast migration.
+TEST fallocate -l 8G $M0/dir1/FILE2
 
 # Rename the file to create a linkto, for rebalance to
 # act on the file
diff --git a/tests/bugs/readdir-ahead/bug-1436090.t b/tests/bugs/readdir-ahead/bug-1436090.t
index 58e9093f1c3..e0877f15684 100755
--- a/tests/bugs/readdir-ahead/bug-1436090.t
+++ b/tests/bugs/readdir-ahead/bug-1436090.t
@@ -19,12 +19,12 @@ EXPECT 'Started' cluster_volinfo_field 1 $V0 'Status';
 TEST glusterfs -s $H1 --volfile-id $V0 $M0;
 TEST mkdir $M0/dir1
 
-# Create a large file (3.2 GB), so that rebalance takes time
-# Reading from /dev/urandom is slow, so we will cat it together
-dd if=/dev/urandom of=/tmp/FILE2 bs=64k count=10240
-for i in {1..5}; do
-  cat /tmp/FILE2 >> $M0/dir1/foo
-done
+# Create a large file (4 GB), so that rebalance takes time
+# Since we really don't care about the contents of the file, we use fallocate
+# to generate the file much faster. We could also use truncate, which is even
+# faster, but rebalance could take advantage of an sparse file and migrate it
+# in an optimized way, but we don't want a fast migration.
+TEST fallocate -l 4G $M0/dir1/foo
 
 TEST mv $M0/dir1/foo $M0/dir1/bar
 
diff --git a/tests/bugs/replicate/bug-1046624.t b/tests/bugs/replicate/bug-1046624.t
index 9ae40879228..e2762ea6764 100755
--- a/tests/bugs/replicate/bug-1046624.t
+++ b/tests/bugs/replicate/bug-1046624.t
@@ -25,11 +25,12 @@ TEST $CLI volume start $V0;
 EXPECT 'Started' volinfo_field $V0 'Status';
 
 ## Mount native
-TEST glusterfs --volfile-server=$H0 --volfile-id=$V0 $M0 --use-readdirp=no
+TEST ${GFS} --volfile-server=$H0 --volfile-id=$V0 --use-readdirp=no $M0
 
 TEST `echo "TEST-FILE" > $M0/File`
 TEST `mkdir $M0/Dir`
 TEST kill_brick $V0 $H0 $B0/${V0}-0
+EXPECT_WITHIN ${PROCESS_DOWN_TIMEOUT} "^0$" afr_child_up_status $V0 0
 
 TEST `ln -s $M0/File $M0/Link1`
 TEST `ln -s $M0/Dir $M0/Link2`
diff --git a/tests/bugs/replicate/bug-1130892.t b/tests/bugs/replicate/bug-1130892.t
index 0f57d669674..c7509f33cc2 100644
--- a/tests/bugs/replicate/bug-1130892.t
+++ b/tests/bugs/replicate/bug-1130892.t
@@ -17,9 +17,9 @@ EXPECT 'Created' volinfo_field $V0 'Status';
 TEST gluster volume set $V0 self-heal-daemon off
 
 # Enable Client side heal
-TEST $CLI volume set $V0 cluster.data-self-heal on
-TEST $CLI volume set $V0 cluster.metadata-self-heal on
-TEST $CLI volume set $V0 cluster.entry-self-heal on
+TEST $CLI volume set $V0 cluster.data-self-heal off
+TEST $CLI volume set $V0 cluster.metadata-self-heal off
+TEST $CLI volume set $V0 cluster.entry-self-heal off
 
 # Disable all perf-xlators
 TEST $CLI volume set $V0 performance.quick-read off
@@ -33,7 +33,7 @@ TEST $CLI volume start $V0;
 EXPECT 'Started' volinfo_field $V0 'Status';
 
 # FUSE Mount
-TEST glusterfs -s $H0 --volfile-id $V0 $M0
+TEST ${GFS} -s $H0 --volfile-id $V0 $M0
 
 # Create files and dirs
 TEST mkdir -p $M0/one/two/
@@ -41,6 +41,7 @@ TEST `echo "Carpe diem" > $M0/one/two/three`
 
 # Simulate disk-replacement
 TEST kill_brick $V0 $H0 $B0/${V0}-1
+EXPECT_WITHIN ${PROCESS_DOWN_TIMEOUT} "^0$" afr_child_up_status $V0 1
 TEST rm -rf $B0/${V0}-1/one
 TEST rm -rf $B0/${V0}-1/.glusterfs
 
@@ -55,10 +56,12 @@ EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status $V0 1
 
 TEST stat $M0/one
 
+sleep 1
+
 # Check pending xattrs
 EXPECT "00000000" afr_get_specific_changelog_xattr $B0/${V0}-0/one trusted.afr.$V0-client-1 data
 EXPECT_NOT "00000000" afr_get_specific_changelog_xattr $B0/${V0}-0/one trusted.afr.$V0-client-1 entry
-EXPECT "00000000" afr_get_specific_changelog_xattr $B0/${V0}-0/one trusted.afr.$V0-client-1 metadata
+EXPECT_NOT "00000000" afr_get_specific_changelog_xattr $B0/${V0}-0/one trusted.afr.$V0-client-1 metadata
 
 TEST gluster volume set $V0 self-heal-daemon on
 
diff --git a/tests/bugs/replicate/bug-1180545.t b/tests/bugs/replicate/bug-1180545.t
index ccf51d8b8a6..5e40edd6c38 100644
--- a/tests/bugs/replicate/bug-1180545.t
+++ b/tests/bugs/replicate/bug-1180545.t
@@ -7,6 +7,31 @@
 . $(dirname $0)/../../volume.rc
 . $(dirname $0)/../../afr.rc
 
+function check_sh_entries() {
+        local expected="$1"
+        local count=
+        local good="0"
+        shift
+
+        for i in $*; do
+                count="$(count_sh_entries $i)"
+                if [[ "x${count}" == "x${expected}" ]]; then
+                        good="$((good + 1))"
+                fi
+        done
+        if [[ "x${good}" != "x${last_good}" ]]; then
+                last_good="${good}"
+# This triggers a sweep of the heal index. However if more than one brick
+# tries to heal the same directory at the same time, one of them will take
+# the lock and the other will give up, waiting for the next heal cycle, which
+# is set to 60 seconds (the minimum valid value). So, each time we detect
+# that one brick has completed the heal, we trigger another heal.
+                $CLI volume heal $V0
+        fi
+
+        echo "${good}"
+}
+
 cleanup;
 
 TEST glusterd
@@ -15,6 +40,7 @@ TEST pidof glusterd
 TEST $CLI volume create $V0 replica 2 $H0:$B0/brick{0,1}
 TEST $CLI volume set $V0 cluster.heal-timeout 60
 TEST $CLI volume set $V0 cluster.self-heal-daemon off
+TEST $CLI volume set $V0 stat-prefetch off
 TEST $CLI volume start $V0
 TEST $GFS --volfile-id=$V0 --volfile-server=$H0 $M0;
 
@@ -38,10 +64,13 @@ TEST $CLI volume set $V0 cluster.self-heal-daemon on
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "Y" glustershd_up_status
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 0
 EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 1
+
 TEST $CLI volume heal $V0
 
-EXPECT_WITHIN $HEAL_TIMEOUT '2' count_sh_entries $B0/brick0
-EXPECT_WITHIN $HEAL_TIMEOUT '2' count_sh_entries $B0/brick1
+last_good=""
+
+EXPECT_WITHIN $HEAL_TIMEOUT "2" check_sh_entries 2 $B0/brick{0,1}
+
 #Two entries for DIR and two for FILE
 EXPECT_WITHIN $HEAL_TIMEOUT "4" get_pending_heal_count $V0
 TEST diff <(ls $B0/brick0/DIR) <(ls $B0/brick1/DIR)
diff --git a/tests/bugs/replicate/bug-1221481-allow-fops-on-dir-split-brain.t b/tests/bugs/replicate/bug-1221481-allow-fops-on-dir-split-brain.t
index c4752c488f4..6ff471fbf15 100644
--- a/tests/bugs/replicate/bug-1221481-allow-fops-on-dir-split-brain.t
+++ b/tests/bugs/replicate/bug-1221481-allow-fops-on-dir-split-brain.t
@@ -11,19 +11,27 @@ TEST pidof glusterd;
 TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1};
 TEST $CLI volume set $V0 cluster.self-heal-daemon off
 TEST $CLI volume start $V0;
-TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0
+TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0
 TEST mkdir $M0/dir
 TEST touch $M0/dir/file{1..5}
 
 #Create entry split-brain
 TEST kill_brick $V0 $H0 $B0/$V0"1"
+EXPECT_WITHIN ${PROCESS_DOWN_TIMEOUT} "^0$" afr_child_up_status $V0 1
 TEST touch $M0/dir/FILE
+EXPECT_WITHIN ${UMOUNT_TIMEOUT} "^Y$" force_umount $M0
 TEST $CLI volume start $V0 force
+TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0
+EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 0
 EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 1
 TEST kill_brick $V0 $H0 $B0/$V0"0"
+EXPECT_WITHIN ${PROCESS_DOWN_TIMEOUT} "^0$" afr_child_up_status $V0 0
 TEST touch $M0/dir/FILE
+EXPECT_WITHIN ${UMOUNT_TIMEOUT} "^Y$" force_umount $M0
 TEST $CLI volume start $V0 force
+TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0
 EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 0
+EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 1
 
 cd $M0/dir
 EXPECT "6" echo $(ls | wc -l)
diff --git a/tests/bugs/replicate/bug-1493415-gfid-heal.t b/tests/bugs/replicate/bug-1493415-gfid-heal.t
index 125c35a7a21..8a79febf4b4 100644
--- a/tests/bugs/replicate/bug-1493415-gfid-heal.t
+++ b/tests/bugs/replicate/bug-1493415-gfid-heal.t
@@ -27,6 +27,11 @@ gfid_str_f1=$(gf_gfid_xattr_to_str $gfid_f1)
 TEST setfattr -x trusted.gfid $B0/${V0}1/f1
 TEST rm $B0/${V0}1/.glusterfs/${gfid_str_f1:0:2}/${gfid_str_f1:2:2}/$gfid_str_f1
 
+# storage/posix considers that a file without gfid changed less than a second
+# before doesn't exist, so we need to wait for a second to force posix to
+# consider that this is a valid file but without gfid.
+sleep 2
+
 # Assume there were no pending xattrs on parent dir due to 1st brick crashing
 # too. Then name heal from client must heal the gfid.
 EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
@@ -52,6 +57,11 @@ TEST rm $B0/${V0}1/.glusterfs/${gfid_str_f2:0:2}/${gfid_str_f2:2:2}/$gfid_str_f2
 TEST setfattr -n trusted.afr.$V0-client-1 -v 0x000000000000000000000001 $B0/${V0}0/dir
 create_brick_xattrop_entry $B0/${V0}0 dir
 
+# storage/posix considers that a file without gfid changed less than a second
+# before doesn't exist, so we need to wait for a second to force posix to
+# consider that this is a valid file but without gfid.
+sleep 2
+
 #Trigger entry-heal via shd
 TEST $CLI volume set $V0 self-heal-daemon on
 EXPECT_WITHIN $PROCESS_UP_TIMEOUT "Y" glustershd_up_status
diff --git a/tests/bugs/replicate/bug-977797.t b/tests/bugs/replicate/bug-977797.t
index c2c0e67ebff..9a8f36c956c 100755
--- a/tests/bugs/replicate/bug-977797.t
+++ b/tests/bugs/replicate/bug-977797.t
@@ -30,7 +30,7 @@ TEST $CLI volume set $V0 cluster.data-self-heal on
 TEST $CLI volume set $V0 cluster.metadata-self-heal on
 TEST $CLI volume set $V0 cluster.entry-self-heal on
 
-TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0
+TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0
 
 
 TEST mkdir -p $M0/a
@@ -77,7 +77,7 @@ afr_get_specific_changelog_xattr $B0/$V0"2"/a/file trusted.afr.$V0-client-1 "dat
 EXPECT_WITHIN $HEAL_TIMEOUT "00000000" \
 afr_get_specific_changelog_xattr $B0/$V0"1"/a trusted.afr.$V0-client-0 "entry"
 
-EXPECT_WITHIN HEAL_TIMEOUT "00000000" \
+EXPECT_WITHIN $HEAL_TIMEOUT "00000000" \
 afr_get_specific_changelog_xattr $B0/$V0"1"/a trusted.afr.$V0-client-1 "entry"
 
 EXPECT_WITHIN $HEAL_TIMEOUT "00000000" \
diff --git a/tests/bugs/snapshot/bug-1109889.t b/tests/bugs/snapshot/bug-1109889.t
index 6b29cdd9eb1..5fdc7dc9506 100644
--- a/tests/bugs/snapshot/bug-1109889.t
+++ b/tests/bugs/snapshot/bug-1109889.t
@@ -19,9 +19,9 @@ TEST $CLI volume create $V0 $H0:$L1 $H0:$L2 $H0:$L3;
 
 TEST $CLI volume start $V0;
 
-TEST glusterfs --volfile-server=$H0 --volfile-id=$V0 $M0;
+TEST $GFS --volfile-server=$H0 --volfile-id=$V0 $M0;
 
-MOUNT_PID=`ps ax |grep "glusterfs --volfile-sever $H0 --volfile-id=$V0 $M0" | grep -v grep | awk '{print $1}' | head -1`
+MOUNT_PID=$(get_mount_process_pid $V0 $M0)
 
 for i in {1..10} ; do echo "file" > $M0/file$i ; done