From 67279d73ac48986a02f657e6eccc157578c22881 Mon Sep 17 00:00:00 2001
From: Richard Wareing <rwareing@fb.com>
Date: Tue, 29 Sep 2015 19:46:14 -0700
Subject: cluster/afr: SHD should always inspect directory heals

Summary:
- This change ensures SHD always inspect directories which are queued
  for healing; i.e. it will not exclusively trust the wise-fool
  algorithm as there are cases where the change log simply isn't
  correct (bugs, crashes, etc).  Failing to perform the entry heal
  in these cases will result in data heals failing to take place.
- We made a similar change in 3.4.x for similar reasons

Test Plan: - Run prove -v tests/basic/shd_force_inspect.t

Reviewers: moox, dph, sshreyas

Reviewed By: sshreyas

Differential Revision: https://phabricator.fb.com/D2492993

Tasks: 8549168

Signature: t1:2492993:1443740894:7cf07168ca09946df9d8f96a3085fe2d3c201543

Change-Id: I2d8e1cbecbbca720cc3ee988d7aae08bea0a5453
---
 tests/basic/shd_force_inspect.t | 60 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 60 insertions(+)
 create mode 100644 tests/basic/shd_force_inspect.t

(limited to 'tests/basic')

diff --git a/tests/basic/shd_force_inspect.t b/tests/basic/shd_force_inspect.t
new file mode 100644
index 00000000000..ecb4fb0d0f4
--- /dev/null
+++ b/tests/basic/shd_force_inspect.t
@@ -0,0 +1,60 @@
+#!/bin/bash
+
+. $(dirname $0)/../include.rc
+. $(dirname $0)/../volume.rc
+
+cleanup;
+
+TEST glusterd
+TEST pidof glusterd
+TEST $CLI volume info;
+
+# Setup a cluster with 3 replicas, and fav child by majority on
+TEST $CLI volume create $V0 replica 3 $H0:$B0/${V0}{1..3};
+TEST $CLI volume set $V0 cluster.choose-local off
+TEST $CLI volume set $V0 cluster.self-heal-daemon on
+TEST $CLI volume set $V0 nfs.disable on
+TEST $CLI volume set $V0 cluster.quorum-type none
+TEST $CLI volume set $V0 cluster.favorite-child-by-majority on
+TEST $CLI volume set $V0 cluster.favorite-child-by-mtime on
+TEST $CLI volume set $V0 cluster.metadata-self-heal off
+TEST $CLI volume set $V0 cluster.data-self-heal off
+TEST $CLI volume set $V0 cluster.entry-self-heal off
+TEST $CLI volume start $V0
+sleep 5
+
+# Part I: FUSE Test
+TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 \
+  --attribute-timeout=0 --entry-timeout=0
+
+cd $M0
+mkdir foo
+dd if=/dev/urandom of=foo/testfile bs=128k count=5 2>/dev/null
+MD5=$(md5sum foo/testfile | cut -d\  -f1)
+
+# Kill the SHD while we setup the test
+pkill -f gluster/glustershd
+
+# Grab the GFID of the file and parent dir
+GFID_PARENT_RAW=$(getfattr -n trusted.gfid -e hex $B0/${V0}1/foo 2>/dev/null | grep trusted.gfid | cut -d= -f2)
+GFID_PARENT_FORMATTED=$(echo "$GFID_PARENT_RAW" | awk '{print substr($1,3,8)"-"substr($1,11,4)"-"substr($1,15,4)"-"substr($1,19,4)"-"substr($1,23,12)}')
+GFID_RAW=$(getfattr -n trusted.gfid -e hex $B0/${V0}1/foo/testfile 2>/dev/null | grep trusted.gfid | cut -d= -f2)
+GFID_FORMATTED=$(echo "$GFID_RAW" | awk '{print substr($1,3,8)"-"substr($1,11,4)"-"substr($1,15,4)"-"substr($1,19,4)"-"substr($1,23,12)}')
+GFID_LINK_B1="$B0/${V0}1/.glusterfs/$(echo $GFID_RAW | awk '{print substr($0,3,2)"/"substr($0,5,2)"/"substr($1,3,8)"-"substr($1,11,4)"-"substr($1,15,4)"-"substr($1,19,4)"-"substr($1,23,12)}')"
+
+# Nuke the file from brick 1
+rm -f $GFID_LINK_B1
+rm -f $B0/${V0}1/foo/testfile
+
+# Now manually queue up the parent directory for healing
+touch $B0/${V0}2/.glusterfs/indices/xattrop/$GFID_PARENT_FORMATTED
+touch $B0/${V0}3/.glusterfs/indices/xattrop/$GFID_PARENT_FORMATTED
+
+# Kick off the SHD and wait 30 seconds for healing to take place
+TEST gluster vol start patchy force
+EXPECT_WITHIN 30 "0" afr_get_pending_heal_count $V0
+
+# Verify the file was healed back to brick 1
+TEST stat $B0/${V0}1/foo/testfile
+
+cleanup
-- 
cgit