afr: fix incorrect reporting of directory split-brain

Problem:
When a directory has dirty xattrs due to failed post-ops or when
replace/reset brick is performed, AFR does a conservative merge as
expected, but heal-info reports it as split-brain because there are no
clear sources.

Fix:
Modify pending flag to contain information about pending heals and
split-brains. For directories, if spit-brain flag is not set,just show
them as needing heal and not being in split-brain.

Fixes: bz#1626994
Change-Id: I09ef821f6887c87d315ae99e6b1de05103cd9383
Signed-off-by: Ravishankar N <ravishankar@redhat.com>

2 files changed, 63 insertions, 1 deletions

diff --git a/tests/afr.rc b/tests/afr.rc
index f253e77561e..35f352df78f 100644
--- a/tests/afr.rc
+++ b/tests/afr.rc
@@ -2,7 +2,7 @@
 
 function create_brick_xattrop_entry {
         local xattrop_dir=$(afr_get_index_path $1)
-        local base_entry=`ls $xattrop_dir`
+        local base_entry=`ls $xattrop_dir|grep xattrop`
         local gfid_str
         local params=`echo "$@" | cut -d' ' -f2-`
         echo $params
diff --git a/tests/bugs/replicate/bug-1626994-info-split-brain.t b/tests/bugs/replicate/bug-1626994-info-split-brain.t
new file mode 100644
index 00000000000..86bfecb1a9e
--- /dev/null
+++ b/tests/bugs/replicate/bug-1626994-info-split-brain.t
@@ -0,0 +1,62 @@
+#!/bin/bash
+
+. $(dirname $0)/../../include.rc
+. $(dirname $0)/../../volume.rc
+. $(dirname $0)/../../afr.rc
+
+cleanup;
+
+# Test to check dirs having dirty xattr do not show up in info split-brain.
+
+TEST glusterd;
+TEST pidof glusterd;
+TEST $CLI volume create $V0 replica 3 $H0:$B0/${V0}{0,1,2};
+TEST $CLI volume set $V0 self-heal-daemon off
+TEST $CLI volume start $V0;
+EXPECT 'Started' volinfo_field $V0 'Status';
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" brick_up_status $V0 $H0 $B0/${V0}0
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" brick_up_status $V0 $H0 $B0/${V0}1
+EXPECT_WITHIN $PROCESS_UP_TIMEOUT "1" brick_up_status $V0 $H0 $B0/${V0}2
+TEST $GFS --volfile-id=/$V0 --volfile-server=$H0 $M0
+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
+
+# Create base entry in indices/xattrop
+echo "Data" > $M0/FILE
+rm -f $M0/FILE
+EXPECT "1" count_index_entries $B0/${V0}0
+EXPECT "1" count_index_entries $B0/${V0}1
+EXPECT "1" count_index_entries $B0/${V0}2
+
+TEST mkdir $M0/dirty_dir
+TEST mkdir $M0/pending_dir
+
+# Set dirty xattrs on all bricks to simulate the case where entry transaction
+# succeeded only the pre-op phase.
+TEST setfattr -n trusted.afr.dirty -v 0x000000000000000000000001 $B0/${V0}0/dirty_dir
+TEST setfattr -n trusted.afr.dirty -v 0x000000000000000000000001 $B0/${V0}1/dirty_dir
+TEST setfattr -n trusted.afr.dirty -v 0x000000000000000000000001 $B0/${V0}2/dirty_dir
+create_brick_xattrop_entry $B0/${V0}0  dirty_dir
+# Should not show up as split-brain.
+EXPECT "0" afr_get_split_brain_count $V0
+
+# replace/reset brick case where the new brick has dirty and the other 2 bricks
+# blame it should not be reported as split-brain.
+TEST setfattr -n trusted.afr.$V0-client-2 -v 0x000000000000000000000001 $B0/${V0}0
+TEST setfattr -n trusted.afr.$V0-client-2 -v 0x000000000000000000000001 $B0/${V0}1
+TEST setfattr -n trusted.afr.dirty -v 0x000000000000000000000001 $B0/${V0}2
+create_brick_xattrop_entry $B0/${V0}0 "/"
+# Should not show up as split-brain.
+EXPECT "0" afr_get_split_brain_count $V0
+
+# Set pending xattrs on all bricks blaming each other to simulate the case of
+# entry split-brain.
+TEST setfattr -n trusted.afr.$V0-client-1 -v 0x000000000000000000000001 $B0/${V0}0/pending_dir
+TEST setfattr -n trusted.afr.$V0-client-2 -v 0x000000000000000000000001 $B0/${V0}1/pending_dir
+TEST setfattr -n trusted.afr.$V0-client-0 -v 0x000000000000000000000001 $B0/${V0}2/pending_dir
+create_brick_xattrop_entry $B0/${V0}0 pending_dir
+# Should show up as split-brain.
+EXPECT "1" afr_get_split_brain_count $V0
+
+cleanup;