From c8be9af5f54aa3e42dffee4024d2915743258242 Mon Sep 17 00:00:00 2001 From: Ravishankar N Date: Thu, 5 Mar 2015 11:29:25 +0000 Subject: afr: remove stale index entries Problem: During pre-op phase, the index xlator 1. Creates the entry inside .glusterfs/indices/xattrop 2. Winds the xattrop fop to posix to mark dirty/pending changelogs. If the brick crashes after 1, the xattrop entry becomes stale and never gets removed by shd during subsequent crawls because there is nothing to heal (changelogs are zero). Though the stale entry does not get displayed in the output of 'heal info' command, it nevertheless stays there forever unless a new write transaction is performed on the file. Fix: During index self-heal if afr xattrs are found to be clean (indicated by ret value of 2 on a call to afr_shd_selfheal(), send a dummy post-op with all 0s for the xattr values, which makes the index xlator to unlink the stale entry. Change-Id: I02cb2bc937f2e3f3f3cb35d67b006664dc7ef919 BUG: 1190069 Signed-off-by: Ravishankar N Reviewed-on: http://review.gluster.org/9714 Tested-by: Gluster Build System Reviewed-by: Anuradha Talur Reviewed-by: Pranith Kumar Karampuri Tested-by: Pranith Kumar Karampuri --- .../bug-1190069-afr-stale-index-entries.t | 57 ++++++++++++++++++++++ 1 file changed, 57 insertions(+) create mode 100644 tests/bugs/replicate/bug-1190069-afr-stale-index-entries.t (limited to 'tests/bugs/replicate/bug-1190069-afr-stale-index-entries.t') diff --git a/tests/bugs/replicate/bug-1190069-afr-stale-index-entries.t b/tests/bugs/replicate/bug-1190069-afr-stale-index-entries.t new file mode 100644 index 0000000000..a476563e31 --- /dev/null +++ b/tests/bugs/replicate/bug-1190069-afr-stale-index-entries.t @@ -0,0 +1,57 @@ +#!/bin/bash +. $(dirname $0)/../../include.rc +. $(dirname $0)/../../volume.rc +. $(dirname $0)/../../afr.rc +cleanup; + +#Stale entries in xattrop folder for files which do not need heal must be removed during the next index heal crawl. + +TEST glusterd; +TEST pidof glusterd; +TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1}; +TEST $CLI volume set $V0 cluster.self-heal-daemon off +TEST $CLI volume start $V0; +TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0 +TEST `echo hello>$M0/datafile` +TEST touch $M0/mdatafile + +#Create split-brain and reset the afr xattrs, so that we have only the entry inside xattrop folder. +#This is to simulate the case where the brick crashed just before pre-op happened, but index xlator created the entry inside xattrop folder. + +#Create data, metadata SB. +TEST kill_brick $V0 $H0 $B0/$V0"1" +TEST stat $M0/datafile +TEST `echo append>>$M0/datafile` +TEST chmod +x $M0/mdatafile +TEST $CLI volume start $V0 force +EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 1 +TEST kill_brick $V0 $H0 $B0/$V0"0" +TEST stat $M0/datafile +TEST `echo append>>$M0/datafile` +TEST chmod +x $M0/mdatafile +TEST $CLI volume start $V0 force +EXPECT_WITHIN $CHILD_UP_TIMEOUT '1' afr_child_up_status_meta $M0 $V0-replicate-0 0 +TEST ! cat $M0/datafile + +TEST $CLI volume set $V0 cluster.self-heal-daemon on +EXPECT_WITHIN $PROCESS_UP_TIMEOUT "Y" glustershd_up_status +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 + +TEST $CLI volume heal $V0 +EXPECT_WITHIN $HEAL_TIMEOUT '2' echo $(count_sh_entries $B0/$V0"0") +EXPECT_WITHIN $HEAL_TIMEOUT '2' echo $(count_sh_entries $B0/$V0"1") + +#Reset xattrs and trigger heal. +TEST setfattr -n trusted.afr.$V0-client-1 -v 0x000000000000000000000000 $B0/${V0}0/datafile +TEST setfattr -n trusted.afr.$V0-client-0 -v 0x000000000000000000000000 $B0/${V0}1/datafile + +TEST setfattr -n trusted.afr.$V0-client-1 -v 0x000000000000000000000000 $B0/${V0}0/mdatafile +TEST setfattr -n trusted.afr.$V0-client-0 -v 0x000000000000000000000000 $B0/${V0}1/mdatafile + +TEST $CLI volume heal $V0 +EXPECT_WITHIN $HEAL_TIMEOUT "0" afr_get_pending_heal_count $V0 +EXPECT_WITHIN $HEAL_TIMEOUT '0' echo $(count_sh_entries $B0/$V0"0") +EXPECT_WITHIN $HEAL_TIMEOUT '0' echo $(count_sh_entries $B0/$V0"1") + +cleanup -- cgit