Ocfs2 devel - Oct 2013 - [PATCH 2/5] Add DLM recovery callbacks

recover_prep() is called when DLM understands a node is down.
recover_slot() is called once all nodes have acknowledged recover_prep
and recovery can begin.
recover_done() is called once the recovery is complete. It returns the
new membership.

Signed-off-by: Goldwyn Rodrigues <rgoldwyn at suse.com>
---
 fs/ocfs2/stack_user.c | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c
index 286edf1..bf3e2fc 100644
--- a/fs/ocfs2/stack_user.c
+++ b/fs/ocfs2/stack_user.c
@@ -110,6 +110,8 @@
 struct ocfs2_live_connection {
 	struct list_head		oc_list;
 	struct ocfs2_cluster_connection	*oc_conn;
+	atomic_t                        oc_this_node;
+	int                             oc_our_slot;
 };
 
 struct ocfs2_control_private {
@@ -799,6 +801,44 @@ static int fs_protocol_compare(struct
ocfs2_protocol_version *existing,
 	return 0;
 }
 
+static void user_recover_prep(void *arg)
+{
+}
+
+static void user_recover_slot(void *arg, struct dlm_slot *slot)
+{
+	struct ocfs2_cluster_connection *conn +		(struct ocfs2_cluster_connection  *)
arg;
+	printk(KERN_INFO "ocfs2: Node %d/%d down. Initiating recovery.\n",
+			slot->nodeid, slot->slot);
+	conn->cc_recovery_handler(slot->nodeid, conn->cc_recovery_data);
+
+}
+
+static void user_recover_done(void *arg, struct dlm_slot *slots,
+		int num_slots, int our_slot,
+		uint32_t generation)
+{
+	struct ocfs2_cluster_connection *conn +		(struct ocfs2_cluster_connection
*)arg;
+	struct ocfs2_live_connection *lc = conn->cc_private;
+	int i;
+
+	for (i = 0; i < num_slots; i++)
+		if (slots[i].slot == our_slot) {
+			atomic_set(&lc->oc_this_node, slots[i].nodeid);
+			break;
+		}
+
+	lc->oc_our_slot = our_slot;
+}
+
+const struct dlm_lockspace_ops ocfs2_ls_ops = {
+	.recover_prep = user_recover_prep,
+	.recover_slot = user_recover_slot,
+	.recover_done = user_recover_done,
+};
+
 static int user_cluster_connect(struct ocfs2_cluster_connection *conn)
 {
 	dlm_lockspace_t *fsdlm;
-- 
1.8.1.4


-- 
Goldwyn

On Thu, 3 Oct 2013 00:49:37 -0500 Goldwyn Rodrigues <rgoldwyn at suse.de>
wrote:
> recover_prep() is called when DLM understands a node is down.
> recover_slot() is called once all nodes have acknowledged recover_prep
> and recovery can begin.
> recover_done() is called once the recovery is complete. It returns the
> new membership.
> 
> ...
>
> --- a/fs/ocfs2/stack_user.c
> +++ b/fs/ocfs2/stack_user.c
> @@ -110,6 +110,8 @@
>  struct ocfs2_live_connection {
>  	struct list_head		oc_list;
>  	struct ocfs2_cluster_connection	*oc_conn;
> +	atomic_t                        oc_this_node;
> +	int                             oc_our_slot;
>  };
>  
>  struct ocfs2_control_private {
> @@ -799,6 +801,44 @@ static int fs_protocol_compare(struct
ocfs2_protocol_version *existing,
>  	return 0;
>  }
>  
> +static void user_recover_prep(void *arg)
> +{
> +}
> +
> +static void user_recover_slot(void *arg, struct dlm_slot *slot)
> +{
> +	struct ocfs2_cluster_connection *conn > +		(struct
ocfs2_cluster_connection  *) arg;
A cast of void* is unneeded.  And it's somewhat undesirable, because if
someone stupidly goes and changes that void* to a long or a
task_struct* or whatever, the cast will suppress the warning which we'd
like to get.

> +	printk(KERN_INFO "ocfs2: Node %d/%d down. Initiating
recovery.\n",
> +			slot->nodeid, slot->slot);
> +	conn->cc_recovery_handler(slot->nodeid, conn->cc_recovery_data);
> +
> +}
> +
> +static void user_recover_done(void *arg, struct dlm_slot *slots,
> +		int num_slots, int our_slot,
> +		uint32_t generation)
> +{
> +	struct ocfs2_cluster_connection *conn > +		(struct
ocfs2_cluster_connection *)arg;
etc.
> +	struct ocfs2_live_connection *lc = conn->cc_private;
> +	int i;
> +
> +	for (i = 0; i < num_slots; i++)
> +		if (slots[i].slot == our_slot) {
> +			atomic_set(&lc->oc_this_node, slots[i].nodeid);
> +			break;
> +		}
> +
> +	lc->oc_our_slot = our_slot;
> +}
> +