From 955bfd567329cf7fe63e9c3b89d333a55e5e9a20 Mon Sep 17 00:00:00 2001 From: Mohit Agrawal Date: Tue, 9 Jun 2020 15:38:12 +0530 Subject: test: Test case brick-mux-validation-in-cluster.t is failing on RHEL-8 Brick process are not properly attached on any cluster node while some volume options are changed on peer node and glusterd is down on that specific node. Solution: At the time of restart glusterd it got a friend update request from a peer node if peer node having some changes on volume.If the brick process is started before received a friend update request in that case brick_mux behavior is not workingproperly. All bricks are attached to the same process even volumes options are not the same. To avoid the issue introduce an atomic flag volpeerupdate and update the value while glusterd has received a friend update request from peer for a specific volume.If volpeerupdate flag is 1 volume is started by glusterd_import_friend_volume synctask Change-Id: I4c026f1e7807ded249153670e6967a2be8d22cb7 Credit: Sanju Rakaonde fixes: #1290 Signed-off-by: Mohit Agrawal --- tests/bugs/glusterd/brick-mux-validation-in-cluster.t | 4 +--- xlators/mgmt/glusterd/src/glusterd-utils.c | 6 ++++-- xlators/mgmt/glusterd/src/glusterd.h | 4 ++++ 3 files changed, 9 insertions(+), 5 deletions(-) diff --git a/tests/bugs/glusterd/brick-mux-validation-in-cluster.t b/tests/bugs/glusterd/brick-mux-validation-in-cluster.t index f088dbb426c..b6af487a791 100644 --- a/tests/bugs/glusterd/brick-mux-validation-in-cluster.t +++ b/tests/bugs/glusterd/brick-mux-validation-in-cluster.t @@ -100,10 +100,8 @@ $CLI_2 volume set $V0 performance.readdir-ahead on $CLI_2 volume set $V1 performance.readdir-ahead on TEST $glusterd_1; +EXPECT_WITHIN $PROBE_TIMEOUT 2 peer_count -sleep 10 - -EXPECT 4 count_brick_processes EXPECT_WITHIN $PROCESS_UP_TIMEOUT 4 count_brick_pids EXPECT_WITHIN $PROCESS_UP_TIMEOUT 0 count_N/A_brick_pids diff --git a/xlators/mgmt/glusterd/src/glusterd-utils.c b/xlators/mgmt/glusterd/src/glusterd-utils.c index 78374a03c99..b941073843d 100644 --- a/xlators/mgmt/glusterd/src/glusterd-utils.c +++ b/xlators/mgmt/glusterd/src/glusterd-utils.c @@ -3626,6 +3626,7 @@ glusterd_compare_friend_volume(dict_t *peer_data, int32_t count, "Version of volume %s differ. local version = %d, " "remote version = %d on peer %s", volinfo->volname, volinfo->version, version, hostname); + GF_ATOMIC_INIT(volinfo->volpeerupdate, 1); *status = GLUSTERD_VOL_COMP_UPDATE_REQ; goto out; } else if (version < volinfo->version) { @@ -4746,7 +4747,7 @@ glusterd_volinfo_stop_stale_bricks(glusterd_volinfo_t *new_volinfo, * brick multiplexing enabled, then stop the brick process */ if (ret || (new_brickinfo->snap_status == -1) || - is_brick_mx_enabled()) { + GF_ATOMIC_GET(old_volinfo->volpeerupdate)) { /*TODO: may need to switch to 'atomic' flavour of * brick_stop, once we make peer rpc program also * synctask enabled*/ @@ -6449,7 +6450,8 @@ glusterd_brick_start(glusterd_volinfo_t *volinfo, * three different triggers for an attempt to start the brick process * due to the quorum handling code in glusterd_friend_sm. */ - if (brickinfo->status == GF_BRICK_STARTING || brickinfo->start_triggered) { + if (brickinfo->status == GF_BRICK_STARTING || brickinfo->start_triggered || + GF_ATOMIC_GET(volinfo->volpeerupdate)) { gf_msg_debug(this->name, 0, "brick %s is already in starting " "phase", diff --git a/xlators/mgmt/glusterd/src/glusterd.h b/xlators/mgmt/glusterd/src/glusterd.h index f2651ebe43c..b43e98ad773 100644 --- a/xlators/mgmt/glusterd/src/glusterd.h +++ b/xlators/mgmt/glusterd/src/glusterd.h @@ -513,6 +513,10 @@ struct glusterd_volinfo_ { * volfile generation code, we are * temporarily appending either "-hot" * or "-cold" */ + gf_atomic_t volpeerupdate; + /* Flag to check about volume has received updates + from peer + */ }; typedef enum gd_snap_status_ { -- cgit