From 2b2eb846c49caba13ab92ec66af20292e7780fc1 Mon Sep 17 00:00:00 2001

From: Ravishankar N <ravishankar@redhat.com>

Date: Tue, 11 Feb 2020 14:34:48 +0530

Subject: [PATCH 410/449] afr: prevent spurious entry heals leading to gfid

 split-brain

Problem:

In a hyperconverged setup with granular-entry-heal enabled, if a file is

recreated while one of the bricks is down, and an index heal is triggered

(with the brick still down), entry-self heal was doing a spurious heal

with just the 2 good bricks. It was doing a post-op leading to removal

of the filename from .glusterfs/indices/entry-changes as well as

erroneous setting of afr xattrs on the parent. When the brick came up,

the xattrs were cleared, resulting in the renamed file not getting

healed and leading to gfid split-brain and EIO on the mount.

Fix:

Proceed with entry heal only when shd can connect to all bricks of the replica,

just like in data and metadata heal.

BUG: 1804164

> Upstream patch:https://review.gluster.org/#/c/glusterfs/+/24109/

> fixes: bz#1801624

> Change-Id: I916ae26ad1fabf259bc6362da52d433b7223b17e

> Signed-off-by: Ravishankar N <ravishankar@redhat.com>

Change-Id: I23f57e543cff1e3f35eb8dbc60a2babfae6838c7

Signed-off-by: Ravishankar N <ravishankar@redhat.com>

Reviewed-on: https://code.engineering.redhat.com/gerrit/202395

Tested-by: RHGS Build Bot <nigelb@redhat.com>

Reviewed-by: Sunil Kumar Heggodu Gopala Acharya <sheggodu@redhat.com>

---

 .../bug-1433571-undo-pending-only-on-up-bricks.t   | 18 ++-----

 tests/bugs/replicate/bug-1801624-entry-heal.t      | 58 ++++++++++++++++++++++

 xlators/cluster/afr/src/afr-common.c               |  4 +-

 xlators/cluster/afr/src/afr-self-heal-common.c     |  8 +--

 xlators/cluster/afr/src/afr-self-heal-entry.c      |  6 +--

 xlators/cluster/afr/src/afr-self-heal-name.c       |  2 +-

 xlators/cluster/afr/src/afr-self-heal.h            |  2 -

 7 files changed, 69 insertions(+), 29 deletions(-)

 create mode 100644 tests/bugs/replicate/bug-1801624-entry-heal.t

diff --git a/tests/bugs/replicate/bug-1433571-undo-pending-only-on-up-bricks.t b/tests/bugs/replicate/bug-1433571-undo-pending-only-on-up-bricks.t

index 0767f47..10ce013 100644

--- a/tests/bugs/replicate/bug-1433571-undo-pending-only-on-up-bricks.t

+++ b/tests/bugs/replicate/bug-1433571-undo-pending-only-on-up-bricks.t

@@ -49,25 +49,15 @@ TEST $CLI volume start $V0 force

 EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status $V0 0

 EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status $V0 2

-#Kill brick 0 and turn on the client side heal and do ls to trigger the heal.

-#The pending xattrs on bricks 1 & 2 should have pending entry on brick 0.

-TEST kill_brick $V0 $H0 $B0/${V0}0

+# We were killing one brick and checking that entry heal does not reset the

+# pending xattrs for the down brick. Now that we need all bricks to be up for

+# entry heal, I'm removing that test from the .t

+

 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 ls $M0

-EXPECT "000000000000000000000001" get_hex_xattr trusted.afr.$V0-client-0 $B0/${V0}1

-EXPECT "000000000000000000000001" get_hex_xattr trusted.afr.$V0-client-0 $B0/${V0}2

-EXPECT_WITHIN $HEAL_TIMEOUT "000000000000000000000000" get_hex_xattr trusted.afr.$V0-client-2 $B0/${V0}1

-EXPECT_WITHIN $HEAL_TIMEOUT "000000000000000000000000" get_hex_xattr trusted.afr.$V0-client-1 $B0/${V0}2

-

-#Bring back all the bricks and trigger the heal again by doing ls. Now the

-#pending xattrs on all the bricks should be 0.

-TEST $CLI volume start $V0 force

-EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status $V0 0

-TEST ls $M0

-

 TEST cat $M0/f1

 TEST cat $M0/f2

 TEST cat $M0/f3

diff --git a/tests/bugs/replicate/bug-1801624-entry-heal.t b/tests/bugs/replicate/bug-1801624-entry-heal.t

new file mode 100644

index 0000000..94b4651

--- /dev/null

+++ b/tests/bugs/replicate/bug-1801624-entry-heal.t

@@ -0,0 +1,58 @@

+#!/bin/bash

+

+. $(dirname $0)/../../include.rc

+. $(dirname $0)/../../volume.rc

+cleanup;

+

+TEST glusterd

+TEST pidof glusterd

+TEST $CLI volume create $V0 replica 3 $H0:$B0/brick{0,1,2}

+TEST $CLI volume set $V0 heal-timeout 5

+TEST $CLI volume start $V0

+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

+EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 2

+TEST $CLI volume heal $V0 granular-entry-heal enable

+

+TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 0

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 1

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 2

+echo "Data">$M0/FILE

+ret=$?

+TEST [ $ret -eq 0 ]

+

+# Re-create the file when a brick is down.

+TEST kill_brick $V0 $H0 $B0/brick1

+TEST rm $M0/FILE

+echo "New Data">$M0/FILE

+ret=$?

+TEST [ $ret -eq 0 ]

+EXPECT_WITHIN $HEAL_TIMEOUT "4" get_pending_heal_count $V0

+

+# Launching index heal must not reset parent dir afr xattrs or remove granular entry indices.

+$CLI volume heal $V0 # CLI will fail but heal is launched anyway.

+TEST sleep 5 # give index heal a chance to do one run.

+brick0_pending=$(get_hex_xattr trusted.afr.$V0-client-1 $B0/brick0/)

+brick2_pending=$(get_hex_xattr trusted.afr.$V0-client-1 $B0/brick2/)

+TEST [ $brick0_pending -eq "000000000000000000000002" ]

+TEST [ $brick2_pending -eq "000000000000000000000002" ]

+EXPECT "FILE" ls $B0/brick0/.glusterfs/indices/entry-changes/00000000-0000-0000-0000-000000000001/

+EXPECT "FILE" ls $B0/brick2/.glusterfs/indices/entry-changes/00000000-0000-0000-0000-000000000001/

+

+TEST $CLI volume start $V0 force

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" brick_up_status $V0 $H0 $B0/brick1

+EXPECT_WITHIN $CHILD_UP_TIMEOUT "1" afr_child_up_status_in_shd $V0 1

+$CLI volume heal $V0

+EXPECT_WITHIN $HEAL_TIMEOUT "0" get_pending_heal_count $V0

+

+# No gfid-split-brain (i.e. EIO) must be seen. Try on fresh mount to avoid cached values.

+EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0

+TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 0

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 1

+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" afr_child_up_status $V0 2

+TEST cat $M0/FILE

+

+EXPECT_WITHIN $UMOUNT_TIMEOUT "Y" force_umount $M0

+cleanup;

diff --git a/xlators/cluster/afr/src/afr-common.c b/xlators/cluster/afr/src/afr-common.c

index 32127c6..5806556 100644

--- a/xlators/cluster/afr/src/afr-common.c

+++ b/xlators/cluster/afr/src/afr-common.c

@@ -6629,7 +6629,7 @@ afr_fav_child_reset_sink_xattrs(void *opaque)

         ret = afr_selfheal_inodelk(heal_frame, this, inode, this->name, 0, 0,

                                    locked_on);

         {

-            if (ret < AFR_SH_MIN_PARTICIPANTS)

+            if (ret < priv->child_count)

                 goto data_unlock;

             ret = __afr_selfheal_data_prepare(

                 heal_frame, this, inode, locked_on, sources, sinks,

@@ -6646,7 +6646,7 @@ afr_fav_child_reset_sink_xattrs(void *opaque)

         ret = afr_selfheal_inodelk(heal_frame, this, inode, this->name,

                                    LLONG_MAX - 1, 0, locked_on);

         {

-            if (ret < AFR_SH_MIN_PARTICIPANTS)

+            if (ret < priv->child_count)

                 goto mdata_unlock;

             ret = __afr_selfheal_metadata_prepare(

                 heal_frame, this, inode, locked_on, sources, sinks,

diff --git a/xlators/cluster/afr/src/afr-self-heal-common.c b/xlators/cluster/afr/src/afr-self-heal-common.c

index 81ef38a..ce1ea50 100644

--- a/xlators/cluster/afr/src/afr-self-heal-common.c

+++ b/xlators/cluster/afr/src/afr-self-heal-common.c

@@ -1575,7 +1575,6 @@ afr_selfheal_find_direction(call_frame_t *frame, xlator_t *this,

     char *accused = NULL;      /* Accused others without any self-accusal */

     char *pending = NULL;      /* Have pending operations on others */

     char *self_accused = NULL; /* Accused itself */

-    int min_participants = -1;

     priv = this->private;

@@ -1599,12 +1598,7 @@ afr_selfheal_find_direction(call_frame_t *frame, xlator_t *this,

         }

     }

-    if (type == AFR_DATA_TRANSACTION || type == AFR_METADATA_TRANSACTION) {

-        min_participants = priv->child_count;

-    } else {

-        min_participants = AFR_SH_MIN_PARTICIPANTS;

-    }

-    if (afr_success_count(replies, priv->child_count) < min_participants) {

+    if (afr_success_count(replies, priv->child_count) < priv->child_count) {

         /* Treat this just like locks not being acquired */

         return -ENOTCONN;

     }

diff --git a/xlators/cluster/afr/src/afr-self-heal-entry.c b/xlators/cluster/afr/src/afr-self-heal-entry.c

index 3ce882e..40be898 100644

--- a/xlators/cluster/afr/src/afr-self-heal-entry.c

+++ b/xlators/cluster/afr/src/afr-self-heal-entry.c

@@ -597,7 +597,7 @@ afr_selfheal_entry_dirent(call_frame_t *frame, xlator_t *this, fd_t *fd,

     ret = afr_selfheal_entrylk(frame, this, fd->inode, this->name, NULL,

                                locked_on);

     {

-        if (ret < AFR_SH_MIN_PARTICIPANTS) {

+        if (ret < priv->child_count) {

             gf_msg_debug(this->name, 0,

                          "%s: Skipping "

                          "entry self-heal as only %d sub-volumes "

@@ -991,7 +991,7 @@ __afr_selfheal_entry(call_frame_t *frame, xlator_t *this, fd_t *fd,

     ret = afr_selfheal_entrylk(frame, this, fd->inode, this->name, NULL,

                                data_lock);

     {

-        if (ret < AFR_SH_MIN_PARTICIPANTS) {

+        if (ret < priv->child_count) {

             gf_msg_debug(this->name, 0,

                          "%s: Skipping "

                          "entry self-heal as only %d sub-volumes could "

@@ -1115,7 +1115,7 @@ afr_selfheal_entry(call_frame_t *frame, xlator_t *this, inode_t *inode)

     ret = afr_selfheal_tie_breaker_entrylk(frame, this, inode, priv->sh_domain,

                                            NULL, locked_on);

     {

-        if (ret < AFR_SH_MIN_PARTICIPANTS) {

+        if (ret < priv->child_count) {

             gf_msg_debug(this->name, 0,

                          "%s: Skipping "

                          "entry self-heal as only %d sub-volumes could "

diff --git a/xlators/cluster/afr/src/afr-self-heal-name.c b/xlators/cluster/afr/src/afr-self-heal-name.c

index 36640b5..7d4f208 100644

--- a/xlators/cluster/afr/src/afr-self-heal-name.c

+++ b/xlators/cluster/afr/src/afr-self-heal-name.c

@@ -514,7 +514,7 @@ afr_selfheal_name_do(call_frame_t *frame, xlator_t *this, inode_t *parent,

     ret = afr_selfheal_entrylk(frame, this, parent, this->name, bname,

                                locked_on);

     {

-        if (ret < AFR_SH_MIN_PARTICIPANTS) {

+        if (ret < priv->child_count) {

             ret = -ENOTCONN;

             goto unlock;

         }

diff --git a/xlators/cluster/afr/src/afr-self-heal.h b/xlators/cluster/afr/src/afr-self-heal.h

index 6555ec5..8234cec 100644

--- a/xlators/cluster/afr/src/afr-self-heal.h

+++ b/xlators/cluster/afr/src/afr-self-heal.h

@@ -11,8 +11,6 @@

 #ifndef _AFR_SELFHEAL_H

 #define _AFR_SELFHEAL_H

-#define AFR_SH_MIN_PARTICIPANTS 2

-

 /* Perform fop on all UP subvolumes and wait for all callbacks to return */

 #define AFR_ONALL(frame, rfn, fop, args...)                                    \

-- 

1.8.3.1