From 4c7460554f2f7c6515b77e70c878fc15d3813237 Mon Sep 17 00:00:00 2001

From: Mohit Agrawal <moagrawal@redhat.com>

Date: Sat, 4 Aug 2018 12:05:03 +0530

Subject: [PATCH 344/351] glusterd: Compare volume_id before start/attach a

 brick

Problem: After reboot a node brick is not coming up because

         fsid comparison is failed before start a brick

Solution: Instead of comparing fsid compare volume_id to

          resolve the same because fsid is changed after

          reboot a node but volume_id persist as a xattr

          on brick_root path at the time of creating a volume.

> Change-Id: Ic289aab1b4ebfd83bbcae8438fee26ae61a0fff4

> fixes: bz#1612418

> (Cherry pick from commit bd8fc26a278697c30537d879ea5402db7ebab577)

> (Reviwed on upstream link https://review.gluster.org/#/c/glusterfs/+/20638/)

Change-Id: Ia183c1d67af662584debcb301171c35be5ac0d9d

BUG: 1612098

Signed-off-by: Mohit Agrawal <moagrawal@redhat.com>

Reviewed-on: https://code.engineering.redhat.com/gerrit/146835

Tested-by: Mohit Agrawal <moagrawa@redhat.com>

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

---

 tests/{basic => bugs/glusterd}/bug-1595320.t |  9 +++---

 xlators/mgmt/glusterd/src/glusterd-utils.c   | 47 ++++++++++++++++------------

 2 files changed, 32 insertions(+), 24 deletions(-)

 rename tests/{basic => bugs/glusterd}/bug-1595320.t (94%)

diff --git a/tests/basic/bug-1595320.t b/tests/bugs/glusterd/bug-1595320.t

similarity index 94%

rename from tests/basic/bug-1595320.t

rename to tests/bugs/glusterd/bug-1595320.t

index 9d856ee..f41df9d 100644

--- a/tests/basic/bug-1595320.t

+++ b/tests/bugs/glusterd/bug-1595320.t

@@ -1,8 +1,8 @@

 #!/bin/bash

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

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

-. $(dirname $0)/../snapshot.rc

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

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

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

 cleanup

@@ -52,7 +52,7 @@ EXPECT 0 count_brick_processes

 # Unmount 3rd brick root from node

 brick_root=$L3

-TEST umount -l $brick_root 2>/dev/null

+_umount_lv 3

 # Start the volume only 2 brick should be start

 TEST $CLI volume start $V0 force

@@ -70,6 +70,7 @@ n=`ls -lrth /proc/$brick_pid/fd | grep -iw $L3 | grep -v ".glusterfs" | wc -l`

 TEST [ $n -eq 0 ]

 # Mount the brick root

+TEST mkdir -p $brick_root

 TEST mount -t xfs -o nouuid  /dev/test_vg_3/brick_lvm $brick_root

 # Replace brick_pid file to test brick_attach code

diff --git a/xlators/mgmt/glusterd/src/glusterd-utils.c b/xlators/mgmt/glusterd/src/glusterd-utils.c

index 6f7c787..1752425 100644

--- a/xlators/mgmt/glusterd/src/glusterd-utils.c

+++ b/xlators/mgmt/glusterd/src/glusterd-utils.c

@@ -5439,6 +5439,12 @@ attach_brick_callback (struct rpc_req *req, struct iovec *iov, int count,

         frame->local = NULL;

         frame->cookie = NULL;

+        if (!iov) {

+              gf_log (frame->this->name, GF_LOG_ERROR, "iov is NULL");

+              ret   = -1;

+              goto out;

+        }

+

         ret = xdr_to_generic (*iov, &rsp, (xdrproc_t)xdr_gf_getspec_rsp);

         if (ret < 0) {

                 gf_log (frame->this->name, GF_LOG_ERROR, "XDR decoding error");

@@ -6073,17 +6079,19 @@ glusterd_brick_start (glusterd_volinfo_t *volinfo,

                       gf_boolean_t wait,

                       gf_boolean_t only_connect)

 {

-        int                     ret   = -1;

+        int                      ret   = -1;

         xlator_t                *this = NULL;

         glusterd_brickinfo_t    *other_brick;

         glusterd_conf_t         *conf = NULL;

-        int32_t                 pid                   = -1;

-        char                    pidfile[PATH_MAX]     = {0};

-        char                    socketpath[PATH_MAX]  = {0};

-        char                    *brickpath            = NULL;

+        int32_t                  pid                   = -1;

+        char                     pidfile[PATH_MAX]     = {0};

+        char                     socketpath[PATH_MAX]  = {0};

+        char                    *brickpath             = NULL;

         glusterd_volinfo_t      *other_vol;

-        struct statvfs           brickstat = {0,};

         gf_boolean_t             is_service_running = _gf_false;

+        uuid_t                   volid                 = {0,};

+        ssize_t                  size                  = -1;

+

         this = THIS;

         GF_ASSERT (this);

@@ -6130,24 +6138,23 @@ glusterd_brick_start (glusterd_volinfo_t *volinfo,

         GLUSTERD_GET_BRICK_PIDFILE (pidfile, volinfo, brickinfo, conf);

-        ret = sys_statvfs (brickinfo->path, &brickstat);

-        if (ret) {

-                gf_msg (this->name, GF_LOG_ERROR,

-                        errno, GD_MSG_BRICKINFO_CREATE_FAIL,

-                        "failed to get statfs() call on brick %s",

-                        brickinfo->path);

+        /* Compare volume-id xattr is helpful to ensure the existence of a brick_root

+           path before the start/attach a brick

+        */

+        size = sys_lgetxattr (brickinfo->path, GF_XATTR_VOL_ID_KEY, volid, 16);

+        if (size != 16) {

+                gf_log (this->name, GF_LOG_ERROR,

+                        "Missing %s extended attribute on brick root (%s),"

+                        " brick is deemed not to be a part of the volume (%s) ",

+                        GF_XATTR_VOL_ID_KEY, brickinfo->path, volinfo->volname);

                 goto out;

         }

-        /* Compare fsid is helpful to ensure the existence of a brick_root

-           path before the start/attach a brick

-        */

-        if (brickinfo->statfs_fsid &&

-            (brickinfo->statfs_fsid != brickstat.f_fsid)) {

+        if (strncmp (uuid_utoa (volinfo->volume_id), uuid_utoa(volid), GF_UUID_BUF_SIZE)) {

                 gf_log (this->name, GF_LOG_ERROR,

-                        "fsid comparison is failed it means Brick root path"

-                        " %s is not created by glusterd, start/attach will also fail",

-                        brickinfo->path);

+                        "Mismatching %s extended attribute on brick root (%s),"

+                        " brick is deemed not to be a part of the volume (%s)",

+                        GF_XATTR_VOL_ID_KEY, brickinfo->path, volinfo->volname);

                 goto out;

         }

         is_service_running = gf_is_service_running (pidfile, &pid;;

-- 

1.8.3.1