summaryrefslogtreecommitdiffstats
path: root/tests/bugs/core
diff options
context:
space:
mode:
Diffstat (limited to 'tests/bugs/core')
-rw-r--r--tests/bugs/core/brick-mux-fd-cleanup.t78
-rwxr-xr-xtests/bugs/core/bug-1402841.t-mt-dir-scan-race.t9
-rw-r--r--tests/bugs/core/bug-1432542-mpx-restart-crash.t27
-rw-r--r--tests/bugs/core/bug-1650403.t113
-rw-r--r--tests/bugs/core/bug-1699025-brick-mux-detach-brick-fd-issue.t33
-rw-r--r--tests/bugs/core/bug-834465.c85
-rwxr-xr-xtests/bugs/core/bug-927616.t2
-rwxr-xr-xtests/bugs/core/io-stats-1322825.t12
-rw-r--r--tests/bugs/core/multiplex-limit-issue-151.t57
9 files changed, 301 insertions, 115 deletions
diff --git a/tests/bugs/core/brick-mux-fd-cleanup.t b/tests/bugs/core/brick-mux-fd-cleanup.t
new file mode 100644
index 00000000000..de11c177b8a
--- /dev/null
+++ b/tests/bugs/core/brick-mux-fd-cleanup.t
@@ -0,0 +1,78 @@
+#!/bin/bash
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../volume.rc
+
+#This .t tests that the fds from client are closed on brick when gluster volume
+#stop is executed in brick-mux setup.
+
+cleanup;
+TEST glusterd
+TEST pidof glusterd
+
+function keep_fd_open {
+#This function has to be run as background job because opening the fd in
+#foreground and running commands is leading to flush calls on these fds
+#which is making it very difficult to create the race where fds will be left
+#open even after the brick dies.
+ exec 5>$M1/a
+ exec 6>$M1/b
+ while [ -f $M0/a ]; do sleep 1; done
+}
+
+function count_open_files {
+ local brick_pid="$1"
+ local pattern="$2"
+ ls -l /proc/$brick_pid/fd | grep -i "$pattern" | wc -l
+}
+
+TEST $CLI volume set all cluster.brick-multiplex on
+TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1}
+TEST $CLI volume create $V1 replica 2 $H0:$B0/${V1}{2,3}
+#Have same configuration on both bricks so that they are multiplexed
+#Delay flush fop for a second
+TEST $CLI volume heal $V0 disable
+TEST $CLI volume heal $V1 disable
+TEST $CLI volume set $V0 delay-gen posix
+TEST $CLI volume set $V0 delay-gen.enable flush
+TEST $CLI volume set $V0 delay-gen.delay-percentage 100
+TEST $CLI volume set $V0 delay-gen.delay-duration 1000000
+TEST $CLI volume set $V1 delay-gen posix
+TEST $CLI volume set $V1 delay-gen.enable flush
+TEST $CLI volume set $V1 delay-gen.delay-percentage 100
+TEST $CLI volume set $V1 delay-gen.delay-duration 1000000
+
+TEST $CLI volume start $V0
+TEST $CLI volume start $V1
+
+TEST $GFS -s $H0 --volfile-id=$V0 --direct-io-mode=enable $M0
+TEST $GFS -s $H0 --volfile-id=$V1 --direct-io-mode=enable $M1
+
+TEST touch $M0/a
+keep_fd_open &
+TEST $CLI volume profile $V1 start
+brick_pid=$(get_brick_pid $V1 $H0 $B0/${V1}2)
+TEST count_open_files $brick_pid "$B0/${V1}2/a"
+TEST count_open_files $brick_pid "$B0/${V1}2/b"
+TEST count_open_files $brick_pid "$B0/${V1}3/a"
+TEST count_open_files $brick_pid "$B0/${V1}3/b"
+
+#If any other flush fops are introduced into the system other than the one at
+#cleanup it interferes with the race, so test for it
+EXPECT "^0$" echo "$($CLI volume profile $V1 info incremental | grep -i flush | wc -l)"
+#Stop the volume
+TEST $CLI volume stop $V1
+
+#Wait for cleanup resources or volume V1
+EXPECT_WITHIN $GRAPH_SWITCH_TIMEOUT "^0$" count_open_files $brick_pid "$B0/${V1}2/a"
+EXPECT_WITHIN $GRAPH_SWITCH_TIMEOUT "^0$" count_open_files $brick_pid "$B0/${V1}2/b"
+EXPECT_WITHIN $GRAPH_SWITCH_TIMEOUT "^0$" count_open_files $brick_pid "$B0/${V1}3/a"
+EXPECT_WITHIN $GRAPH_SWITCH_TIMEOUT "^0$" count_open_files $brick_pid "$B0/${V1}3/b"
+
+TEST rm -f $M0/a #Exit keep_fd_open()
+wait
+
+EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0
+EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M1
+
+cleanup
diff --git a/tests/bugs/core/bug-1402841.t-mt-dir-scan-race.t b/tests/bugs/core/bug-1402841.t-mt-dir-scan-race.t
index 6351ba22511..a1b9a851bf7 100755
--- a/tests/bugs/core/bug-1402841.t-mt-dir-scan-race.t
+++ b/tests/bugs/core/bug-1402841.t-mt-dir-scan-race.t
@@ -3,6 +3,8 @@
. $(dirname $0)/../../volume.rc
cleanup;
+FILE_COUNT=500
+
TEST glusterd
TEST pidof glusterd
TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1}
@@ -11,15 +13,14 @@ TEST $CLI volume set $V0 cluster.shd-wait-qlength 100
TEST $CLI volume start $V0
TEST glusterfs --volfile-id=$V0 --volfile-server=$H0 $M0;
-touch $M0/file{1..200}
-
+for i in `seq 1 $FILE_COUNT`; do touch $M0/file$i; done
TEST kill_brick $V0 $H0 $B0/${V0}1
-for i in {1..200}; do echo hello>$M0/file$i; done
+for i in `seq 1 $FILE_COUNT`; do echo hello>$M0/file$i; chmod -x $M0/file$i; done
TEST $CLI volume start $V0 force
EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" brick_up_status $V0 $H0 $B0/${V0}1
EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 1
-EXPECT "200" get_pending_heal_count $V0
+EXPECT "$FILE_COUNT" get_pending_heal_count $V0
TEST $CLI volume set $V0 self-heal-daemon on
EXPECT_WITHIN $PROCESS_UP_TIMEOUT "Y" glustershd_up_status
diff --git a/tests/bugs/core/bug-1432542-mpx-restart-crash.t b/tests/bugs/core/bug-1432542-mpx-restart-crash.t
index 2179630202d..2793d7008e1 100644
--- a/tests/bugs/core/bug-1432542-mpx-restart-crash.t
+++ b/tests/bugs/core/bug-1432542-mpx-restart-crash.t
@@ -1,12 +1,14 @@
#!/bin/bash
-SCRIPT_TIMEOUT=300
+SCRIPT_TIMEOUT=800
. $(dirname $0)/../../include.rc
. $(dirname $0)/../../volume.rc
. $(dirname $0)/../../traps.rc
-NUM_VOLS=20
+cleanup;
+
+NUM_VOLS=15
MOUNT_BASE=$(dirname $M0)
# GlusterD reports that bricks are started when in fact their attach requests
@@ -40,7 +42,7 @@ create_volume () {
local vol_name=$(printf "%s-vol%02d" $V0 $1)
local brick_base=$(get_brick_base $1)
- local cmd="$CLI volume create $vol_name replica 2"
+ local cmd="$CLI volume create $vol_name replica 3"
local b
for b in $(seq 0 5); do
local this_brick=${brick_base}/brick$b
@@ -50,7 +52,7 @@ create_volume () {
TEST $cmd
TEST $CLI volume start $vol_name
# check for 6 bricks and 1 shd daemon to be up and running
- EXPECT_WITHIN $PROCESS_UP_TIMEOUT 7 count_up_bricks $vol_name
+ EXPECT_WITHIN 120 7 count_up_bricks $vol_name
local mount_point=$(get_mount_point $1)
mkdir -p $mount_point
TEST $GFS -s $H0 --volfile-id=$vol_name $mount_point
@@ -77,12 +79,27 @@ TEST $CLI volume set all cluster.brick-multiplex on
# Our infrastructure can't handle an arithmetic expression here. The formula
# is (NUM_VOLS-1)*5 because it sees each TEST/EXPECT once but needs the other
# NUM_VOLS-1 and there are 5 such statements in each iteration.
-TESTS_EXPECTED_IN_LOOP=95
+TESTS_EXPECTED_IN_LOOP=84
for i in $(seq 1 $NUM_VOLS); do
+ starttime="$(date +%s)";
+
create_volume $i
TEST dd if=/dev/zero of=$(get_mount_point $i)/a_file bs=4k count=1
+ # Unmounting to reduce memory footprint on regression hosts
+ mnt_point=$(get_mount_point $i)
+ EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $mnt_point
+ endtime=$(expr $(date +%s) - $starttime)
+
+ echo "Memory Used after $i volumes : $(pmap -x $(pgrep glusterfsd) | grep total)"
+ echo "Thread Count after $i volumes: $(ps -T -p $(pgrep glusterfsd) | wc -l)"
+ echo "Time taken : ${endtime} seconds"
done
+echo "=========="
+echo "List of all the threads in the Brick process"
+ps -T -p $(pgrep glusterfsd)
+echo "=========="
+
# Kill glusterd, and wait a bit for all traces to disappear.
TEST killall -9 glusterd
sleep 5
diff --git a/tests/bugs/core/bug-1650403.t b/tests/bugs/core/bug-1650403.t
new file mode 100644
index 00000000000..43d09bc8bd9
--- /dev/null
+++ b/tests/bugs/core/bug-1650403.t
@@ -0,0 +1,113 @@
+#!/bin/bash
+
+SCRIPT_TIMEOUT=500
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../volume.rc
+. $(dirname $0)/../../traps.rc
+
+cleanup;
+
+NUM_VOLS=5
+MOUNT_BASE=$(dirname $M0)
+
+# GlusterD reports that bricks are started when in fact their attach requests
+# might still need to be retried. That's a bit of a hack, but there's no
+# feasible way to wait at that point (in attach_brick) and the rest of the
+# code is unprepared to deal with transient errors so the whole "brick start"
+# would fail. Meanwhile, glusterfsd can only handle attach requests at a
+# rather slow rate. After GlusterD tries to start a couple of hundred bricks,
+# glusterfsd can fall behind and we start getting mount failures. Arguably,
+# those are spurious because we will eventually catch up. We're just not
+# ready *yet*. More to the point, even if the errors aren't spurious that's
+# not what we're testing right now. Therefore, we give glusterfsd a bit more
+# breathing room for this test than we would otherwise.
+MOUNT_TIMEOUT=15
+
+get_brick_base () {
+ printf "%s/vol%02d" $B0 $1
+}
+
+get_mount_point () {
+ printf "%s/vol%02d" $MOUNT_BASE $1
+}
+
+function count_up_bricks {
+ vol=$1;
+ $CLI --xml volume status $vol | grep '<status>1' | wc -l
+}
+
+create_volume () {
+
+ local vol_name=$(printf "%s-vol%02d" $V0 $1)
+
+ local brick_base=$(get_brick_base $1)
+ local cmd="$CLI volume create $vol_name replica 3"
+ local b
+ for b in $(seq 0 5); do
+ local this_brick=${brick_base}/brick$b
+ mkdir -p $this_brick
+ cmd="$cmd $H0:$this_brick"
+ done
+ TEST $cmd
+ TEST $CLI volume start $vol_name
+ # check for 6 bricks and 1 shd daemon to be up and running
+ EXPECT_WITHIN 120 7 count_up_bricks $vol_name
+ local mount_point=$(get_mount_point $1)
+ mkdir -p $mount_point
+ TEST $GFS -s $H0 --volfile-id=$vol_name $mount_point
+}
+
+cleanup_func () {
+ local v
+ for v in $(seq 1 $NUM_VOLS); do
+ local mount_point=$(get_mount_point $v)
+ force_umount $mount_point
+ rm -rf $mount_point
+ local vol_name=$(printf "%s-vol%02d" $V0 $v)
+ $CLI volume stop $vol_name
+ $CLI volume delete $vol_name
+ rm -rf $(get_brick_base $1) &
+ done &> /dev/null
+ wait
+}
+push_trapfunc cleanup_func
+
+TEST glusterd
+TEST $CLI volume set all cluster.brick-multiplex on
+
+# Our infrastructure can't handle an arithmetic expression here. The formula
+# is (NUM_VOLS-1)*5 because it sees each TEST/EXPECT once but needs the other
+# NUM_VOLS-1 and there are 5 such statements in each iteration.
+TESTS_EXPECTED_IN_LOOP=24
+for i in $(seq 1 $NUM_VOLS); do
+ create_volume $i
+ TEST dd if=/dev/zero of=$(get_mount_point $i)/a_file bs=4k count=1
+ # Unmounting to reduce memory footprint on regression hosts
+ mnt_point=$(get_mount_point $i)
+ EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $mnt_point
+done
+
+glustershd_pid=`ps auxwww | grep glustershd | grep -v grep | awk -F " " '{print $2}'`
+TEST [ $glustershd_pid != 0 ]
+start=`pmap -x $glustershd_pid | grep total | awk -F " " '{print $4}'`
+echo "Memory consumption for glustershd process"
+for i in $(seq 1 50); do
+ pmap -x $glustershd_pid | grep total
+ for j in $(seq 1 $NUM_VOLS); do
+ vol_name=$(printf "%s-vol%02d" $V0 $j)
+ gluster v set $vol_name cluster.self-heal-daemon off > /dev/null
+ gluster v set $vol_name cluster.self-heal-daemon on > /dev/null
+ done
+done
+
+end=`pmap -x $glustershd_pid | grep total | awk -F " " '{print $4}'`
+diff=$((end-start))
+
+# If memory consumption is more than 10M it means some leak in reconfigure
+# code path
+
+TEST [ $diff -lt 10000 ]
+
+trap - EXIT
+cleanup
diff --git a/tests/bugs/core/bug-1699025-brick-mux-detach-brick-fd-issue.t b/tests/bugs/core/bug-1699025-brick-mux-detach-brick-fd-issue.t
new file mode 100644
index 00000000000..1acbaa8dc0b
--- /dev/null
+++ b/tests/bugs/core/bug-1699025-brick-mux-detach-brick-fd-issue.t
@@ -0,0 +1,33 @@
+#!/bin/bash
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../volume.rc
+. $(dirname $0)/../../cluster.rc
+
+function count_brick_processes {
+ pgrep glusterfsd | wc -l
+}
+
+cleanup
+
+#bug-1444596 - validating brick mux
+
+TEST glusterd
+TEST $CLI volume create $V0 $H0:$B0/brick{0,1}
+TEST $CLI volume create $V1 $H0:$B0/brick{2,3}
+
+TEST $CLI volume set all cluster.brick-multiplex on
+
+TEST $CLI volume start $V0
+TEST $CLI volume start $V1
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT 4 online_brick_count
+EXPECT 1 count_brick_processes
+
+TEST $CLI volume stop $V1
+# At the time initialize brick daemon it always keeps open
+# standard fd's (0, 1 , 2) so after stop 1 volume fd's should
+# be open
+nofds=$(ls -lrth /proc/`pgrep glusterfsd`/fd | grep dev/null | wc -l)
+TEST [ $((nofds)) -eq 3 ]
+
+cleanup
diff --git a/tests/bugs/core/bug-834465.c b/tests/bugs/core/bug-834465.c
index 61d3deac077..33dd270b112 100644
--- a/tests/bugs/core/bug-834465.c
+++ b/tests/bugs/core/bug-834465.c
@@ -7,55 +7,54 @@
#include <fcntl.h>
int
-main (int argc, char *argv[])
+main(int argc, char *argv[])
{
- int fd = -1;
- char *filename = NULL;
- struct flock lock = {0, };
- int i = 0;
- int ret = -1;
-
- if (argc != 2) {
- fprintf (stderr, "Usage: %s <filename> ", argv[0]);
- goto out;
+ int fd = -1;
+ char *filename = NULL;
+ struct flock lock = {
+ 0,
+ };
+ int i = 0;
+ int ret = -1;
+
+ if (argc != 2) {
+ fprintf(stderr, "Usage: %s <filename> ", argv[0]);
+ goto out;
+ }
+
+ filename = argv[1];
+
+ fd = open(filename, O_RDWR | O_CREAT, 0);
+ if (fd < 0) {
+ fprintf(stderr, "open (%s) failed (%s)\n", filename, strerror(errno));
+ goto out;
+ }
+
+ lock.l_type = F_WRLCK;
+ lock.l_whence = SEEK_SET;
+ lock.l_start = 1;
+ lock.l_len = 1;
+
+ while (i < 100) {
+ lock.l_type = F_WRLCK;
+ ret = fcntl(fd, F_SETLK, &lock);
+ if (ret < 0) {
+ fprintf(stderr, "fcntl setlk failed (%s)\n", strerror(errno));
+ goto out;
}
- filename = argv[1];
-
- fd = open (filename, O_RDWR | O_CREAT, 0);
- if (fd < 0) {
- fprintf (stderr, "open (%s) failed (%s)\n", filename,
- strerror (errno));
- goto out;
+ lock.l_type = F_UNLCK;
+ ret = fcntl(fd, F_SETLK, &lock);
+ if (ret < 0) {
+ fprintf(stderr, "fcntl setlk failed (%s)\n", strerror(errno));
+ goto out;
}
- lock.l_type = F_WRLCK;
- lock.l_whence = SEEK_SET;
- lock.l_start = 1;
- lock.l_len = 1;
-
- while (i < 100) {
- lock.l_type = F_WRLCK;
- ret = fcntl (fd, F_SETLK, &lock);
- if (ret < 0) {
- fprintf (stderr, "fcntl setlk failed (%s)\n",
- strerror (errno));
- goto out;
- }
-
- lock.l_type = F_UNLCK;
- ret = fcntl (fd, F_SETLK, &lock);
- if (ret < 0) {
- fprintf (stderr, "fcntl setlk failed (%s)\n",
- strerror (errno));
- goto out;
- }
-
- i++;
- }
+ i++;
+ }
- ret = 0;
+ ret = 0;
out:
- return ret;
+ return ret;
}
diff --git a/tests/bugs/core/bug-927616.t b/tests/bugs/core/bug-927616.t
index 6bb64743183..18257131ac7 100755
--- a/tests/bugs/core/bug-927616.t
+++ b/tests/bugs/core/bug-927616.t
@@ -3,6 +3,8 @@
. $(dirname $0)/../../include.rc
. $(dirname $0)/../../nfs.rc
+#G_TESTDEF_TEST_STATUS_CENTOS6=NFS_TEST
+
cleanup;
TEST glusterd
diff --git a/tests/bugs/core/io-stats-1322825.t b/tests/bugs/core/io-stats-1322825.t
index d232ecb2420..53f2d040daa 100755
--- a/tests/bugs/core/io-stats-1322825.t
+++ b/tests/bugs/core/io-stats-1322825.t
@@ -23,7 +23,7 @@ TEST $CLI volume profile $V0 start
TEST mkdir $M0/dir1
# Generate the stat dump across the io-stat instances
-TEST setfattr -n trusted.io-stats-dump -v /tmp/io-stats-1322825 $M0
+TEST setfattr -n trusted.io-stats-dump -v io-stats-1322825 $M0
# Check if $M0 is clean w.r.t xattr information
# TODO: if there are better ways to check we really get no attr error, please
@@ -42,12 +42,12 @@ ret=$(echo $?)
EXPECT 0 echo $ret
# Check if we have 5 io-stat files in /tmp
-EXPECT 5 ls -1 /tmp/io-stats-1322825*
+EXPECT 5 ls -1 /var/run/gluster/io-stats-1322825*
# Cleanup the 5 generated files
-rm -f /tmp/io-stats-1322825*
+rm -f /var/run/gluster/io-stats-1322825*
# Rinse and repeat above for a directory
-TEST setfattr -n trusted.io-stats-dump -v /tmp/io-stats-1322825 $M0/dir1
+TEST setfattr -n trusted.io-stats-dump -v io-stats-1322825 $M0/dir1
getfattr -n trusted.io-stats-dump $B0/${V0}1/dir1 2>&1 | grep -qi "no such attribute"
ret=$(echo $?)
EXPECT 0 echo $ret
@@ -61,7 +61,7 @@ getfattr -n trusted.io-stats-dump $B0/${V0}4/dir1 2>&1 | grep -qi "no such attri
ret=$(echo $?)
EXPECT 0 echo $ret
-EXPECT 5 ls -1 /tmp/io-stats-1322825*
-rm -f /tmp/io-stats-1322825*
+EXPECT 5 ls -1 /var/run/gluster/io-stats-1322825*
+rm -f /var/run/gluster/io-stats-1322825*
cleanup;
diff --git a/tests/bugs/core/multiplex-limit-issue-151.t b/tests/bugs/core/multiplex-limit-issue-151.t
deleted file mode 100644
index c5bbbdad852..00000000000
--- a/tests/bugs/core/multiplex-limit-issue-151.t
+++ /dev/null
@@ -1,57 +0,0 @@
-#!/bin/bash
-
-. $(dirname $0)/../../include.rc
-. $(dirname $0)/../../traps.rc
-. $(dirname $0)/../../volume.rc
-
-function count_up_bricks {
- $CLI --xml volume status all | grep '<status>1' | wc -l
-}
-
-function count_brick_processes {
- pgrep glusterfsd | wc -l
-}
-
-function count_brick_pids {
- $CLI --xml volume status all | sed -n '/.*<pid>\([^<]*\).*/s//\1/p' \
- | grep -v "N/A" | sort | uniq | wc -l
-}
-
-cleanup;
-
-TEST glusterd
-
-TEST $CLI volume set all cluster.brick-multiplex on
-TEST ! $CLI volume set all cluster.max-bricks-per-process -1
-TEST ! $CLI volume set all cluster.max-bricks-per-process foobar
-TEST $CLI volume set all cluster.max-bricks-per-process 3
-
-push_trapfunc "$CLI volume set all cluster.brick-multiplex off"
-push_trapfunc "cleanup"
-
-TEST $CLI volume create $V0 $H0:$B0/brick{0..5}
-TEST $CLI volume start $V0
-
-EXPECT 2 count_brick_processes
-EXPECT 2 count_brick_pids
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 6 count_up_bricks
-
-pkill gluster
-TEST glusterd
-
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 2 count_brick_processes
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 2 count_brick_pids
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 6 count_up_bricks
-
-TEST $CLI volume add-brick $V0 $H0:$B0/brick6
-
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 3 count_brick_processes
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 3 count_brick_pids
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 7 count_up_bricks
-
-TEST $CLI volume remove-brick $V0 $H0:$B0/brick3 start
-TEST $CLI volume remove-brick $V0 $H0:$B0/brick3 force
-
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 3 count_brick_processes
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 3 count_brick_pids
-EXPECT_WITHIN $PROCESS_UP_TIMEOUT 6 count_up_bricks
generated by cgit (git 2.34.1) at 2026-01-05 11:03:42 +0000