From ecaa0f10820f4b6e803021919ce59a43aedf356b Mon Sep 17 00:00:00 2001

From: Ravishankar N <ravishankar@redhat.com>

Date: Thu, 4 Jun 2020 16:15:35 +0530

Subject: [PATCH 402/449] afr: wake up index healer threads

...whenever shd is re-enabled after disabling or there is a change in

`cluster.heal-timeout`, without needing to restart shd or waiting for the

current `cluster.heal-timeout` seconds to expire.

> Upstream patch link:https://review.gluster.org/#/c/glusterfs/+/23288/

> Change-Id: Ia5ebd7c8e9f5b54cba3199c141fdd1af2f9b9bfe

> fixes: bz#1744548

> Reported-by: Glen Kiessling <glenk1973@hotmail.com>

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

BUG: 1764091

Change-Id: I42aa0807f09b5a09510fe9efb4a1697dad3410a3

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

Reviewed-on: https://code.engineering.redhat.com/gerrit/202368

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

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

---

 tests/bugs/replicate/bug-1744548-heal-timeout.t | 42 +++++++++++++++++++++++++

 xlators/cluster/afr/src/afr-common.c            |  6 ++--

 xlators/cluster/afr/src/afr-self-heald.c        | 14 ++++++---

 xlators/cluster/afr/src/afr-self-heald.h        |  3 --

 xlators/cluster/afr/src/afr.c                   | 10 ++++++

 xlators/cluster/afr/src/afr.h                   |  2 ++

 6 files changed, 66 insertions(+), 11 deletions(-)

 create mode 100644 tests/bugs/replicate/bug-1744548-heal-timeout.t

diff --git a/tests/bugs/replicate/bug-1744548-heal-timeout.t b/tests/bugs/replicate/bug-1744548-heal-timeout.t

new file mode 100644

index 0000000..3cb73bc

--- /dev/null

+++ b/tests/bugs/replicate/bug-1744548-heal-timeout.t

@@ -0,0 +1,42 @@

+#!/bin/bash

+

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

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

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

+

+cleanup;

+

+TEST glusterd;

+TEST pidof glusterd;

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

+TEST $CLI volume heal $V0 disable

+TEST $CLI volume start $V0

+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 ! $CLI volume heal $V0

+

+# Enable shd and verify that index crawl is triggered immediately.

+TEST $CLI volume profile $V0 start

+TEST $CLI volume profile $V0 info clear

+TEST $CLI volume heal $V0 enable

+TEST $CLI volume heal $V0

+# Each brick does 3 opendirs, corresponding to dirty, xattrop and entry-changes

+COUNT=`$CLI volume profile $V0 info incremental |grep OPENDIR|awk '{print $8}'|tr -d '\n'`

+TEST [ "$COUNT" == "333" ]

+

+# Check that a change in heal-timeout is honoured immediately.

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

+sleep 10

+COUNT=`$CLI volume profile $V0 info incremental |grep OPENDIR|awk '{print $8}'|tr -d '\n'`

+# Two crawls must have happened.

+TEST [ "$COUNT" == "666" ]

+

+# shd must not heal if it is disabled and heal-timeout is changed.

+TEST $CLI volume heal $V0 disable

+TEST $CLI volume profile $V0 info clear

+TEST $CLI volume set $V0 cluster.heal-timeout 6

+sleep 6

+COUNT=`$CLI volume profile $V0 info incremental |grep OPENDIR|awk '{print $8}'|tr -d '\n'`

+TEST [ -z $COUNT ]

+cleanup;

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

index 3690b84..eef7fd2 100644

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

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

@@ -5613,10 +5613,8 @@ afr_notify(xlator_t *this, int32_t event, void *data, void *data2)

          * b) Already heard from everyone, but we now got a child-up

          *    event.

          */

-        if (have_heard_from_all && priv->shd.iamshd) {

-            for (i = 0; i < priv->child_count; i++)

-                if (priv->child_up[i])

-                    afr_selfheal_childup(this, i);

+        if (have_heard_from_all) {

+            afr_selfheal_childup(this, priv);

         }

     }

 out:

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

index 7eb1207..95ac5f2 100644

--- a/xlators/cluster/afr/src/afr-self-heald.c

+++ b/xlators/cluster/afr/src/afr-self-heald.c

@@ -1258,12 +1258,18 @@ out:

     return ret;

 }

-int

-afr_selfheal_childup(xlator_t *this, int subvol)

+void

+afr_selfheal_childup(xlator_t *this, afr_private_t *priv)

 {

-    afr_shd_index_healer_spawn(this, subvol);

+    int subvol = 0;

-    return 0;

+    if (!priv->shd.iamshd)

+        return;

+    for (subvol = 0; subvol < priv->child_count; subvol++)

+        if (priv->child_up[subvol])

+            afr_shd_index_healer_spawn(this, subvol);

+

+    return;

 }

 int

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

index 7de7c43..1990539 100644

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

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

@@ -60,9 +60,6 @@ typedef struct {

 } afr_self_heald_t;

 int

-afr_selfheal_childup(xlator_t *this, int subvol);

-

-int

 afr_selfheal_daemon_init(xlator_t *this);

 int

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

index 33258a0..8f9e71f 100644

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

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

@@ -141,6 +141,7 @@ reconfigure(xlator_t *this, dict_t *options)

     afr_private_t *priv = NULL;

     xlator_t *read_subvol = NULL;

     int read_subvol_index = -1;

+    int timeout_old = 0;

     int ret = -1;

     int index = -1;

     char *qtype = NULL;

@@ -150,6 +151,7 @@ reconfigure(xlator_t *this, dict_t *options)

     char *locking_scheme = NULL;

     gf_boolean_t consistent_io = _gf_false;

     gf_boolean_t choose_local_old = _gf_false;

+    gf_boolean_t enabled_old = _gf_false;

     priv = this->private;

@@ -255,11 +257,13 @@ reconfigure(xlator_t *this, dict_t *options)

     GF_OPTION_RECONF("ensure-durability", priv->ensure_durability, options,

                      bool, out);

+    enabled_old = priv->shd.enabled;

     GF_OPTION_RECONF("self-heal-daemon", priv->shd.enabled, options, bool, out);

     GF_OPTION_RECONF("iam-self-heal-daemon", priv->shd.iamshd, options, bool,

                      out);

+    timeout_old = priv->shd.timeout;

     GF_OPTION_RECONF("heal-timeout", priv->shd.timeout, options, int32, out);

     GF_OPTION_RECONF("consistent-metadata", priv->consistent_metadata, options,

@@ -283,6 +287,12 @@ reconfigure(xlator_t *this, dict_t *options)

         consistent_io = _gf_false;

     priv->consistent_io = consistent_io;

+    if (priv->shd.enabled) {

+        if ((priv->shd.enabled != enabled_old) ||

+            (timeout_old != priv->shd.timeout))

+            afr_selfheal_childup(this, priv);

+    }

+

     ret = 0;

 out:

     return ret;

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

index e731cfa..18f1a6a 100644

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

+++ b/xlators/cluster/afr/src/afr.h

@@ -1332,4 +1332,6 @@ afr_lookup_has_quorum(call_frame_t *frame, xlator_t *this,

 void

 afr_mark_new_entry_changelog(call_frame_t *frame, xlator_t *this);

+void

+afr_selfheal_childup(xlator_t *this, afr_private_t *priv);

 #endif /* __AFR_H__ */

-- 

1.8.3.1