From fb064ec4e302e59aca9ba8a8d97e4cc2d82d31ef Mon Sep 17 00:00:00 2001
From: Pranith Kumar K <pkarampu@redhat.com>
Date: Mon, 17 Jun 2013 13:06:23 +0530
Subject: nfs: Remove afr split-brain handling in nfs

We added this code as an interim fix until afr can
handle split-brains even when opens are not issued.
Afr code has matured to reject fd based fops when
there are split-brains so we can remove it.

Change-Id: Ib337f78eccee86469a5eaabed1a547a2cea2bdcf
BUG: 974972
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5227
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
---
 tests/bugs/bug-974972.t              | 36 ++++++++++++++++++++++++++++++++++++
 tests/volume.rc                      | 12 ++++++++++++
 xlators/cluster/afr/src/afr-common.c |  5 -----
 xlators/nfs/server/src/nfs-fops.c    | 13 -------------
 4 files changed, 48 insertions(+), 18 deletions(-)
 create mode 100755 tests/bugs/bug-974972.t

diff --git a/tests/bugs/bug-974972.t b/tests/bugs/bug-974972.t
new file mode 100755
index 00000000..15deac09
--- /dev/null
+++ b/tests/bugs/bug-974972.t
@@ -0,0 +1,36 @@
+#!/bin/bash
+
+. $(dirname $0)/../include.rc
+. $(dirname $0)/../volume.rc
+
+#This script checks that nfs mount does not fail lookup on files with split-brain
+cleanup;
+
+TEST glusterd
+TEST pidof glusterd
+TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1}
+TEST $CLI volume set $V0 self-heal-daemon off
+TEST $CLI volume start $V0
+sleep 5
+TEST mount -t nfs -o vers=3 $H0:/$V0 $N0
+TEST touch $N0/1
+TEST kill_brick ${V0} ${H0} ${B0}/${V0}1
+echo abc > $N0/1
+TEST $CLI volume start $V0 force
+EXPECT_WITHIN 20 "Y" nfs_up_status
+EXPECT_WITHIN 20 "1" afr_child_up_status_in_nfs $V0 0
+EXPECT_WITHIN 20 "1" afr_child_up_status_in_nfs $V0 1
+
+TEST kill_brick ${V0} ${H0} ${B0}/${V0}0
+echo def > $N0/1
+TEST $CLI volume start $V0 force
+EXPECT_WITHIN 20 "Y" nfs_up_status
+EXPECT_WITHIN 20 "1" afr_child_up_status_in_nfs $V0 0
+EXPECT_WITHIN 20 "1" afr_child_up_status_in_nfs $V0 1
+
+#Lookup should not fail
+TEST ls $N0/1
+TEST ! cat $N0/1
+
+TEST umount $N0
+cleanup
diff --git a/tests/volume.rc b/tests/volume.rc
index 54b6173c..5bb974e1 100644
--- a/tests/volume.rc
+++ b/tests/volume.rc
@@ -105,6 +105,11 @@ function generate_shd_statedump {
         generate_statedump $(get_shd_process_pid $vol)
 }
 
+function generate_nfs_statedump {
+        local vol=$1
+        generate_statedump $(get_nfs_pid $vol)
+}
+
 function generate_brick_statedump {
         local vol=$1
         local host=$2
@@ -119,6 +124,13 @@ function afr_child_up_status_in_shd {
         _afr_child_up_status $vol $brick_id generate_shd_statedump
 }
 
+function afr_child_up_status_in_nfs {
+        local vol=$1
+        #brick_id is (brick-num in volume info - 1)
+        local brick_id=$2
+        _afr_child_up_status $vol $brick_id generate_nfs_statedump
+}
+
 function nfs_up_status {
         gluster volume status | grep "NFS Server" | awk '{print $6}'
 }
diff --git a/xlators/cluster/afr/src/afr-common.c b/xlators/cluster/afr/src/afr-common.c
index 20513d38..3fcf53d8 100644
--- a/xlators/cluster/afr/src/afr-common.c
+++ b/xlators/cluster/afr/src/afr-common.c
@@ -1632,7 +1632,6 @@ afr_self_heal_lookup_unwind (call_frame_t *frame, xlator_t *this,
         afr_local_t *local = NULL;
         int         ret    = -1;
         dict_t      *xattr = NULL;
-        int32_t     spb    = 0;
 
         local = frame->local;
 
@@ -1663,10 +1662,6 @@ afr_self_heal_lookup_unwind (call_frame_t *frame, xlator_t *this,
                                        local->loc.path,
                                        local->self_heal.actual_sh_started);
                 }
-
-                if (local->loc.inode)
-                        spb = afr_is_split_brain (this, local->loc.inode);
-                ret = dict_set_int32 (xattr, "split-brain", spb);
         }
 out:
         AFR_STACK_UNWIND (lookup, frame, local->op_ret, local->op_errno,
diff --git a/xlators/nfs/server/src/nfs-fops.c b/xlators/nfs/server/src/nfs-fops.c
index b6edc99c..fa8f7319 100644
--- a/xlators/nfs/server/src/nfs-fops.c
+++ b/xlators/nfs/server/src/nfs-fops.c
@@ -385,19 +385,6 @@ nfs_fop_lookup_cbk (call_frame_t *frame, void *cookie, xlator_t *this,
 {
         struct nfs_fop_local    *local = NULL;
         fop_lookup_cbk_t        progcbk;
-        int32_t                 spb = 0;
-
-        /*
-         * With native protocol, self-heal failures would be detected during
-         * open.  NFS doesn't issue that open when revalidating cache, so we
-         * have to check for failures here instead.
-         */
-        if (dict_get_int32(xattr, "split-brain", &spb) == 0) {
-                if (spb) {
-                        op_ret = -1;
-                        op_errno = EIO;
-                }
-        }
 
         if (op_ret == 0) {
                 nfs_fix_generation(this,inode);
-- 
cgit