[Ocfs2-devel] [patch 08/11] ocfs2/dlm: wait for dlm recovery done when migrating all lockres
Changwei Ge
ge.changwei at h3c.com
Thu Nov 30 17:59:29 PST 2017
I NACK to this patch since I don't think it can solve the issue Jun
reported completely.
Thanks,
Changwei
On 2017/12/1 6:26, akpm at linux-foundation.org wrote:
> From: piaojun <piaojun at huawei.com>
> Subject: ocfs2/dlm: wait for dlm recovery done when migrating all lockres
>
> wait for dlm recovery done when migrating all lockres in case of new
> lockres to be left after leaving dlm domain.
>
> NodeA NodeB NodeC
>
> umount and migrate
> all lockres
>
> node down
>
> do recovery for NodeB
> and collect a new lockres
> form other live nodes
>
> leave domain but the
> new lockres remains
>
> mount and join domain
>
> request for the owner
> of the new lockres, but
> all the other nodes said
> 'NO', so NodeC decide to
> the owner, and send do
> assert msg to other nodes.
>
> other nodes receive the msg
> and found two masters exist.
> at last cause BUG in
> dlm_assert_master_handler()
> -->BUG();
>
> Link: https://urldefense.proofpoint.com/v2/url?u=http-3A__lkml.kernel.org_r_59FFB7AD.90108-40huawei.com&d=DwICAg&c=RoP1YumCXCgaWHvlZYR8PZh8Bv7qIrMUB65eapI_JnE&r=C7gAd4uDxlAvTdc0vmU6X8CMk6L2iDY8-HD0qT6Fo7Y&m=eA-mkBvutvIzGPMzaE9P8Cis0buQmhfPdaUQEjK-634&s=zTKHn7nr2VuSBip021NkdyIlj7eSYEoKose-sPG4gA0&e=
> Fixes: bc9838c4d44a ("dlm: allow dlm do recovery during shutdown")
> Signed-off-by: Jun Piao <piaojun at huawei.com>
> Reviewed-by: Alex Chen <alex.chen at huawei.com>
> Reviewed-by: Yiwen Jiang <jiangyiwen at huawei.com>
> Cc: Mark Fasheh <mfasheh at versity.com>
> Cc: Joel Becker <jlbec at evilplan.org>
> Cc: Junxiao Bi <junxiao.bi at oracle.com>
> Cc: Joseph Qi <jiangqi903 at gmail.com>
> Signed-off-by: Andrew Morton <akpm at linux-foundation.org>
> ---
>
> fs/ocfs2/dlm/dlmcommon.h | 1 +
> fs/ocfs2/dlm/dlmdomain.c | 14 ++++++++++++++
> fs/ocfs2/dlm/dlmrecovery.c | 13 ++++++++++---
> 3 files changed, 25 insertions(+), 3 deletions(-)
>
> diff -puN fs/ocfs2/dlm/dlmcommon.h~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres fs/ocfs2/dlm/dlmcommon.h
> --- a/fs/ocfs2/dlm/dlmcommon.h~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres
> +++ a/fs/ocfs2/dlm/dlmcommon.h
> @@ -140,6 +140,7 @@ struct dlm_ctxt
> u8 node_num;
> u32 key;
> u8 joining_node;
> + u8 migrate_done; /* set to 1 means node has migrated all lockres */
> wait_queue_head_t dlm_join_events;
> unsigned long live_nodes_map[BITS_TO_LONGS(O2NM_MAX_NODES)];
> unsigned long domain_map[BITS_TO_LONGS(O2NM_MAX_NODES)];
> diff -puN fs/ocfs2/dlm/dlmdomain.c~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres fs/ocfs2/dlm/dlmdomain.c
> --- a/fs/ocfs2/dlm/dlmdomain.c~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres
> +++ a/fs/ocfs2/dlm/dlmdomain.c
> @@ -461,6 +461,18 @@ redo_bucket:
> cond_resched_lock(&dlm->spinlock);
> num += n;
> }
> +
> + if (!num) {
> + if (dlm->reco.state & DLM_RECO_STATE_ACTIVE) {
> + mlog(0, "%s: perhaps there are more lock resources need to "
> + "be migrated after dlm recovery\n", dlm->name);
> + ret = -EAGAIN;
> + } else {
> + mlog(0, "%s: we won't do dlm recovery after migrating all lockres",
> + dlm->name);
> + dlm->migrate_done = 1;
> + }
> + }
> spin_unlock(&dlm->spinlock);
> wake_up(&dlm->dlm_thread_wq);
>
> @@ -2052,6 +2064,8 @@ static struct dlm_ctxt *dlm_alloc_ctxt(c
> dlm->joining_node = DLM_LOCK_RES_OWNER_UNKNOWN;
> init_waitqueue_head(&dlm->dlm_join_events);
>
> + dlm->migrate_done = 0;
> +
> dlm->reco.new_master = O2NM_INVALID_NODE_NUM;
> dlm->reco.dead_node = O2NM_INVALID_NODE_NUM;
>
> diff -puN fs/ocfs2/dlm/dlmrecovery.c~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres fs/ocfs2/dlm/dlmrecovery.c
> --- a/fs/ocfs2/dlm/dlmrecovery.c~ocfs2-dlm-wait-for-dlm-recovery-done-when-migrating-all-lockres
> +++ a/fs/ocfs2/dlm/dlmrecovery.c
> @@ -423,12 +423,11 @@ void dlm_wait_for_recovery(struct dlm_ct
>
> static void dlm_begin_recovery(struct dlm_ctxt *dlm)
> {
> - spin_lock(&dlm->spinlock);
> + assert_spin_locked(&dlm->spinlock);
> BUG_ON(dlm->reco.state & DLM_RECO_STATE_ACTIVE);
> printk(KERN_NOTICE "o2dlm: Begin recovery on domain %s for node %u\n",
> dlm->name, dlm->reco.dead_node);
> dlm->reco.state |= DLM_RECO_STATE_ACTIVE;
> - spin_unlock(&dlm->spinlock);
> }
>
> static void dlm_end_recovery(struct dlm_ctxt *dlm)
> @@ -456,6 +455,13 @@ static int dlm_do_recovery(struct dlm_ct
>
> spin_lock(&dlm->spinlock);
>
> + if (dlm->migrate_done) {
> + mlog(0, "%s: no need do recovery after migrating all lockres\n",
> + dlm->name);
> + spin_unlock(&dlm->spinlock);
> + return 0;
> + }
> +
> /* check to see if the new master has died */
> if (dlm->reco.new_master != O2NM_INVALID_NODE_NUM &&
> test_bit(dlm->reco.new_master, dlm->recovery_map)) {
> @@ -490,12 +496,13 @@ static int dlm_do_recovery(struct dlm_ct
> mlog(0, "%s(%d):recovery thread found node %u in the recovery map!\n",
> dlm->name, task_pid_nr(dlm->dlm_reco_thread_task),
> dlm->reco.dead_node);
> - spin_unlock(&dlm->spinlock);
>
> /* take write barrier */
> /* (stops the list reshuffling thread, proxy ast handling) */
> dlm_begin_recovery(dlm);
>
> + spin_unlock(&dlm->spinlock);
> +
> if (dlm->reco.new_master == dlm->node_num)
> goto master_here;
>
> _
>
> _______________________________________________
> Ocfs2-devel mailing list
> Ocfs2-devel at oss.oracle.com
> https://oss.oracle.com/mailman/listinfo/ocfs2-devel
>
More information about the Ocfs2-devel
mailing list