Attachment 15862 Details for Bug 14295 – Patch for 4.11 cherry-picked from master (after some preliminary cosmetic cherry-picks)

[patch] Patch for 4.11 cherry-picked from master (after some preliminary cosmetic cherry-picks)

bug14295-v411.patch (text/plain), 30.02 KB, created by Martin Schwenke on 2020-03-14 02:54:39 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Martin Schwenke

Created: 2020-03-14 02:54:39 UTC

Size: 30.02 KB

patch

obsolete

>From f3692f8cc68f00b985655159e8235dd0a8448abc Mon Sep 17 00:00:00 2001
>From: Martin Schwenke <martin@meltin.net>
>Date: Tue, 12 Nov 2019 12:04:22 +1100
>Subject: [PATCH 01/11] ctdb-daemon: Rename ctdb_node private_data to
> transport_data
>
>This gives a casual reader a useful clue.
>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>Reviewed-by: Amitay Isaacs <amitay@gmail.com>
>(cherry picked from commit 53f8492caafa8556d0c2d3f272d08ce5ce098c25)
>---
> ctdb/ib/ibw_ctdb.c          | 11 ++++++++---
> ctdb/ib/ibw_ctdb_init.c     |  5 +++--
> ctdb/include/ctdb_private.h |  2 +-
> ctdb/tcp/tcp_connect.c      | 10 +++++-----
> ctdb/tcp/tcp_init.c         |  8 ++++----
> ctdb/tcp/tcp_io.c           |  4 ++--
> 6 files changed, 23 insertions(+), 17 deletions(-)
>
>diff --git a/ctdb/ib/ibw_ctdb.c b/ctdb/ib/ibw_ctdb.c
>index 458646faae0..53911240ff7 100644
>--- a/ctdb/ib/ibw_ctdb.c
>+++ b/ctdb/ib/ibw_ctdb.c
>@@ -55,7 +55,8 @@ int ctdb_ibw_get_address(struct ctdb_context *ctdb,
> 
> int ctdb_ibw_node_connect(struct ctdb_node *node)
> {
>-	struct ctdb_ibw_node *cn = talloc_get_type(node->private_data, struct ctdb_ibw_node);
>+	struct ctdb_ibw_node *cn = talloc_get_type(node->transport_data,
>+						   struct ctdb_ibw_node);
> 	int	rc;
> 
> 	assert(cn!=NULL);
>@@ -118,7 +119,9 @@ int ctdb_ibw_connstate_handler(struct ibw_ctx *ctx, struct ibw_conn *conn)
> 		case IBWC_CONNECTED: { /* after ibw_accept or ibw_connect */
> 			struct ctdb_node *node = talloc_get_type(conn->conn_userdata, struct ctdb_node);
> 			if (node!=NULL) { /* after ibw_connect */
>-				struct ctdb_ibw_node *cn = talloc_get_type(node->private_data, struct ctdb_ibw_node);
>+				struct ctdb_ibw_node *cn = talloc_get_type(
>+					node->transport_data,
>+					struct ctdb_ibw_node);
> 
> 				node->ctdb->upcalls->node_connected(node);
> 				ctdb_flush_cn_queue(cn);
>@@ -136,7 +139,9 @@ int ctdb_ibw_connstate_handler(struct ibw_ctx *ctx, struct ibw_conn *conn)
> 		case IBWC_ERROR: {
> 			struct ctdb_node *node = talloc_get_type(conn->conn_userdata, struct ctdb_node);
> 			if (node!=NULL) {
>-				struct ctdb_ibw_node *cn = talloc_get_type(node->private_data, struct ctdb_ibw_node);
>+				struct ctdb_ibw_node *cn = talloc_get_type(
>+					node->transport_data,
>+					struct ctdb_ibw_node);
> 				struct ibw_ctx *ictx = cn->conn->ctx;
> 
> 				DEBUG(DEBUG_DEBUG, ("IBWC_ERROR, reconnecting...\n"));
>diff --git a/ctdb/ib/ibw_ctdb_init.c b/ctdb/ib/ibw_ctdb_init.c
>index 7e77ec08031..6fdb0d887cf 100644
>--- a/ctdb/ib/ibw_ctdb_init.c
>+++ b/ctdb/ib/ibw_ctdb_init.c
>@@ -67,7 +67,7 @@ static int ctdb_ibw_add_node(struct ctdb_node *node)
> 
> 	assert(cn!=NULL);
> 	cn->conn = ibw_conn_new(ictx, node);
>-	node->private_data = (void *)cn;
>+	node->transport_data = (void *)cn;
> 
> 	return (cn->conn!=NULL ? 0 : -1);
> }
>@@ -153,7 +153,8 @@ int ctdb_flush_cn_queue(struct ctdb_ibw_node *cn)
> 
> static int ctdb_ibw_queue_pkt(struct ctdb_node *node, uint8_t *data, uint32_t length)
> {
>-	struct ctdb_ibw_node *cn = talloc_get_type(node->private_data, struct ctdb_ibw_node);
>+	struct ctdb_ibw_node *cn = talloc_get_type(node->transport_data,
>+						   struct ctdb_ibw_node);
> 	int	rc;
> 
> 	assert(length>=sizeof(uint32_t));
>diff --git a/ctdb/include/ctdb_private.h b/ctdb/include/ctdb_private.h
>index 1f168dae2b8..e9e4d6a7f06 100644
>--- a/ctdb/include/ctdb_private.h
>+++ b/ctdb/include/ctdb_private.h
>@@ -74,7 +74,7 @@ struct ctdb_node {
> 	struct ctdb_context *ctdb;
> 	ctdb_sock_addr address;
> 	const char *name; /* for debug messages */
>-	void *private_data; /* private to transport */
>+	void *transport_data; /* private to transport */
> 	uint32_t pnn;
> 	uint32_t flags;
> 
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 04897f44249..2261d730d52 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -43,7 +43,7 @@
> void ctdb_tcp_stop_connection(struct ctdb_node *node)
> {
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
>-		node->private_data, struct ctdb_tcp_node);
>+		node->transport_data, struct ctdb_tcp_node);
> 
> 	TALLOC_FREE(tnode->out_queue);
> 	TALLOC_FREE(tnode->connect_te);
>@@ -63,7 +63,7 @@ void ctdb_tcp_tnode_cb(uint8_t *data, size_t cnt, void *private_data)
> {
> 	struct ctdb_node *node = talloc_get_type(private_data, struct ctdb_node);
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
>-		node->private_data, struct ctdb_tcp_node);
>+		node->transport_data, struct ctdb_tcp_node);
> 
> 	if (data == NULL) {
> 		node->ctdb->upcalls->node_dead(node);
>@@ -85,7 +85,7 @@ static void ctdb_node_connect_write(struct tevent_context *ev,
> {
> 	struct ctdb_node *node = talloc_get_type(private_data,
> 						 struct ctdb_node);
>-	struct ctdb_tcp_node *tnode = talloc_get_type(node->private_data,
>+	struct ctdb_tcp_node *tnode = talloc_get_type(node->transport_data,
> 						      struct ctdb_tcp_node);
> 	struct ctdb_context *ctdb = node->ctdb;
> 	int error = 0;
>@@ -167,7 +167,7 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
> {
> 	struct ctdb_node *node = talloc_get_type(private_data,
> 						 struct ctdb_node);
>-	struct ctdb_tcp_node *tnode = talloc_get_type(node->private_data, 
>+	struct ctdb_tcp_node *tnode = talloc_get_type(node->transport_data,
> 						      struct ctdb_tcp_node);
> 	struct ctdb_context *ctdb = node->ctdb;
>         ctdb_sock_addr sock_in;
>@@ -300,7 +300,7 @@ static void ctdb_listen_event(struct tevent_context *ev, struct tevent_fd *fde,
> 		return;
> 	}
> 
>-	tnode = talloc_get_type_abort(node->private_data,
>+	tnode = talloc_get_type_abort(node->transport_data,
> 				      struct ctdb_tcp_node);
> 	if (tnode == NULL) {
> 		/* This can't happen - see ctdb_tcp_initialise() */
>diff --git a/ctdb/tcp/tcp_init.c b/ctdb/tcp/tcp_init.c
>index 0eb9799ac4a..9ad13aabc74 100644
>--- a/ctdb/tcp/tcp_init.c
>+++ b/ctdb/tcp/tcp_init.c
>@@ -58,7 +58,7 @@ static int ctdb_tcp_add_node(struct ctdb_node *node)
> 	tnode->out_fd = -1;
> 	tnode->ctdb = node->ctdb;
> 
>-	node->private_data = tnode;
>+	node->transport_data = tnode;
> 	talloc_set_destructor(tnode, tnode_destructor);
> 
> 	return 0;
>@@ -97,7 +97,7 @@ static int ctdb_tcp_connect_node(struct ctdb_node *node)
> {
> 	struct ctdb_context *ctdb = node->ctdb;
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
>-		node->private_data, struct ctdb_tcp_node);
>+		node->transport_data, struct ctdb_tcp_node);
> 
> 	/* startup connection to the other server - will happen on
> 	   next event loop */
>@@ -118,7 +118,7 @@ static int ctdb_tcp_connect_node(struct ctdb_node *node)
> static void ctdb_tcp_restart(struct ctdb_node *node)
> {
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
>-		node->private_data, struct ctdb_tcp_node);
>+		node->transport_data, struct ctdb_tcp_node);
> 
> 	DEBUG(DEBUG_NOTICE,("Tearing down connection to dead node :%d\n", node->pnn));
> 
>@@ -143,7 +143,7 @@ static void ctdb_tcp_shutdown(struct ctdb_context *ctdb)
> 	ctdb->private_data = NULL;
> 
> 	for (i=0; i<ctdb->num_nodes; i++) {
>-		TALLOC_FREE(ctdb->nodes[i]->private_data);
>+		TALLOC_FREE(ctdb->nodes[i]->transport_data);
> 	}
> }
> 
>diff --git a/ctdb/tcp/tcp_io.c b/ctdb/tcp/tcp_io.c
>index 2d8ec0f7062..df9ca02b413 100644
>--- a/ctdb/tcp/tcp_io.c
>+++ b/ctdb/tcp/tcp_io.c
>@@ -39,7 +39,7 @@ void ctdb_tcp_read_cb(uint8_t *data, size_t cnt, void *args)
> {
> 	struct ctdb_node *node = talloc_get_type_abort(args, struct ctdb_node);
> 	struct ctdb_tcp_node *tnode = talloc_get_type_abort(
>-		node->private_data, struct ctdb_tcp_node);
>+		node->transport_data, struct ctdb_tcp_node);
> 	struct ctdb_req_header *hdr = (struct ctdb_req_header *)data;
> 
> 	if (data == NULL) {
>@@ -86,7 +86,7 @@ failed:
> */
> int ctdb_tcp_queue_pkt(struct ctdb_node *node, uint8_t *data, uint32_t length)
> {
>-	struct ctdb_tcp_node *tnode = talloc_get_type(node->private_data,
>+	struct ctdb_tcp_node *tnode = talloc_get_type(node->transport_data,
> 						      struct ctdb_tcp_node);
> 	if (tnode->out_queue == NULL) {
> 		DBG_DEBUG("No outgoing connection, dropping packet\n");
>-- 
>2.25.1
>
>
>From 8c47a57a988d6ae38009394bc04fff588d606964 Mon Sep 17 00:00:00 2001
>From: Martin Schwenke <martin@meltin.net>
>Date: Tue, 12 Nov 2019 12:12:46 +1100
>Subject: [PATCH 02/11] ctdb-daemon: Rename ctdb_context private_data to
> transport_data
>
>This gives a casual reader a useful clue.
>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>Reviewed-by: Amitay Isaacs <amitay@gmail.com>
>(cherry picked from commit 750f3938e4fcd6743954db6b1132751a90ee6107)
>---
> ctdb/ib/ibw_ctdb_init.c     | 8 +++++---
> ctdb/include/ctdb_private.h | 2 +-
> ctdb/tcp/tcp_connect.c      | 7 ++++---
> ctdb/tcp/tcp_init.c         | 8 ++++----
> 4 files changed, 14 insertions(+), 11 deletions(-)
>
>diff --git a/ctdb/ib/ibw_ctdb_init.c b/ctdb/ib/ibw_ctdb_init.c
>index 6fdb0d887cf..f9d00c60605 100644
>--- a/ctdb/ib/ibw_ctdb_init.c
>+++ b/ctdb/ib/ibw_ctdb_init.c
>@@ -40,7 +40,8 @@
> 
> static int ctdb_ibw_listen(struct ctdb_context *ctdb, int backlog)
> {
>-	struct ibw_ctx *ictx = talloc_get_type(ctdb->private_data, struct ibw_ctx);
>+	struct ibw_ctx *ictx = talloc_get_type(ctdb->transport_data,
>+					       struct ibw_ctx);
> 
> 	assert(ictx!=NULL);
> 
>@@ -62,7 +63,8 @@ static int ctdb_ibw_listen(struct ctdb_context *ctdb, int backlog)
>  */
> static int ctdb_ibw_add_node(struct ctdb_node *node)
> {
>-	struct ibw_ctx *ictx = talloc_get_type(node->ctdb->private_data, struct ibw_ctx);
>+	struct ibw_ctx *ictx = talloc_get_type(node->ctdb->transport_data,
>+					       struct ibw_ctx);
> 	struct ctdb_ibw_node *cn = talloc_zero(node, struct ctdb_ibw_node);
> 
> 	assert(cn!=NULL);
>@@ -246,7 +248,7 @@ int ctdb_ibw_init(struct ctdb_context *ctdb)
> 	}
> 
> 	ctdb->methods = &ctdb_ibw_methods;
>-	ctdb->private_data = ictx;
>+	ctdb->transport_data = ictx;
> 	
> 	DEBUG(DEBUG_DEBUG, ("ctdb_ibw_init succeeded.\n"));
> 	return 0;
>diff --git a/ctdb/include/ctdb_private.h b/ctdb/include/ctdb_private.h
>index e9e4d6a7f06..7f160c0c9db 100644
>--- a/ctdb/include/ctdb_private.h
>+++ b/ctdb/include/ctdb_private.h
>@@ -286,7 +286,7 @@ struct ctdb_context {
> 	char *err_msg;
> 	const struct ctdb_methods *methods; /* transport methods */
> 	const struct ctdb_upcalls *upcalls; /* transport upcalls */
>-	void *private_data; /* private to transport */
>+	void *transport_data; /* private to transport */
> 	struct ctdb_db_context *db_list;
> 	struct srvid_context *srv;
> 	struct srvid_context *tunnels;
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 2261d730d52..3e9cf062184 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -277,7 +277,8 @@ static void ctdb_listen_event(struct tevent_context *ev, struct tevent_fd *fde,
> 			      uint16_t flags, void *private_data)
> {
> 	struct ctdb_context *ctdb = talloc_get_type(private_data, struct ctdb_context);
>-	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->private_data, struct ctdb_tcp);
>+	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
>+						struct ctdb_tcp);
> 	ctdb_sock_addr addr;
> 	socklen_t len;
> 	int fd;
>@@ -368,7 +369,7 @@ static void ctdb_listen_event(struct tevent_context *ev, struct tevent_fd *fde,
> */
> static int ctdb_tcp_listen_automatic(struct ctdb_context *ctdb)
> {
>-	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->private_data,
>+	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
> 						struct ctdb_tcp);
>         ctdb_sock_addr sock;
> 	int lock_fd;
>@@ -509,7 +510,7 @@ failed:
> */
> int ctdb_tcp_listen(struct ctdb_context *ctdb)
> {
>-	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->private_data,
>+	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
> 						struct ctdb_tcp);
>         ctdb_sock_addr sock;
> 	int sock_size;
>diff --git a/ctdb/tcp/tcp_init.c b/ctdb/tcp/tcp_init.c
>index 9ad13aabc74..5bd9e2f3a09 100644
>--- a/ctdb/tcp/tcp_init.c
>+++ b/ctdb/tcp/tcp_init.c
>@@ -135,12 +135,12 @@ static void ctdb_tcp_restart(struct ctdb_node *node)
> */
> static void ctdb_tcp_shutdown(struct ctdb_context *ctdb)
> {
>-	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->private_data,
>+	struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
> 						struct ctdb_tcp);
> 	uint32_t i;
> 
> 	talloc_free(ctcp);
>-	ctdb->private_data = NULL;
>+	ctdb->transport_data = NULL;
> 
> 	for (i=0; i<ctdb->num_nodes; i++) {
> 		TALLOC_FREE(ctdb->nodes[i]->transport_data);
>@@ -191,7 +191,7 @@ static const struct ctdb_methods ctdb_tcp_methods = {
> 
> static int tcp_ctcp_destructor(struct ctdb_tcp *ctcp)
> {
>-	ctcp->ctdb->private_data = NULL;
>+	ctcp->ctdb->transport_data = NULL;
> 	ctcp->ctdb->methods = NULL;
> 	
> 	return 0;
>@@ -209,7 +209,7 @@ int ctdb_tcp_init(struct ctdb_context *ctdb)
> 
> 	ctcp->listen_fd = -1;
> 	ctcp->ctdb      = ctdb;
>-	ctdb->private_data = ctcp;
>+	ctdb->transport_data = ctcp;
> 	ctdb->methods = &ctdb_tcp_methods;
> 
> 	talloc_set_destructor(ctcp, tcp_ctcp_destructor);
>-- 
>2.25.1
>
>
>From dca87ea1693f63192b4c23ba97b847ce976a937b Mon Sep 17 00:00:00 2001
>From: Ralph Boehme <slow@samba.org>
>Date: Sat, 29 Feb 2020 12:26:19 +0100
>Subject: [PATCH 03/11] ctdb-daemon: ensure restart() callback is called in
> half-connected state
>
>If NODE_FLAGS_DISCONNECTED is set the node can be in half-connected state. With
>this change we ensure to restart the transport for this case.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Ralph Boehme <slow@samba.org>
>Reviewed-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit 6a4fa0785fc83561939fa41617d526eb96c1af89)
>---
> ctdb/server/ctdb_server.c | 13 ++++++-------
> 1 file changed, 6 insertions(+), 7 deletions(-)
>
>diff --git a/ctdb/server/ctdb_server.c b/ctdb/server/ctdb_server.c
>index 9724d1fe0a8..c5d92af0cfb 100644
>--- a/ctdb/server/ctdb_server.c
>+++ b/ctdb/server/ctdb_server.c
>@@ -301,6 +301,12 @@ done:
> */
> void ctdb_node_dead(struct ctdb_node *node)
> {
>+	if (node->ctdb->methods == NULL) {
>+		DEBUG(DEBUG_ERR,(__location__ " Can not restart transport while shutting down daemon.\n"));
>+		return;
>+	}
>+
>+	node->ctdb->methods->restart(node);
> 	if (node->flags & NODE_FLAGS_DISCONNECTED) {
> 		DEBUG(DEBUG_INFO,("%s: node %s is already marked disconnected: %u connected\n", 
> 			 node->ctdb->name, node->name, 
>@@ -315,13 +321,6 @@ void ctdb_node_dead(struct ctdb_node *node)
> 	DEBUG(DEBUG_ERR,("%s: node %s is dead: %u connected\n",
> 		 node->ctdb->name, node->name, node->ctdb->num_connected));
> 	ctdb_daemon_cancel_controls(node->ctdb, node);
>-
>-	if (node->ctdb->methods == NULL) {
>-		DEBUG(DEBUG_ERR,(__location__ " Can not restart transport while shutting down daemon.\n"));
>-		return;
>-	}
>-
>-	node->ctdb->methods->restart(node);
> }
> 
> /*
>-- 
>2.25.1
>
>
>From 39add06579a648b47b4c51729f2b1a0e6362c40b Mon Sep 17 00:00:00 2001
>From: Martin Schwenke <martin@meltin.net>
>Date: Sun, 1 Mar 2020 16:40:41 +1100
>Subject: [PATCH 04/11] ctdb-daemon: more logical whitespace, debug
> modernisation
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>Reviewed-by: Ralph Boehme <slow@samba.org>
>(cherry picked from commit 15762a34559599cf908e30651a2d4c11560068ed)
>---
> ctdb/server/ctdb_server.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
>diff --git a/ctdb/server/ctdb_server.c b/ctdb/server/ctdb_server.c
>index c5d92af0cfb..4b4c2e9896f 100644
>--- a/ctdb/server/ctdb_server.c
>+++ b/ctdb/server/ctdb_server.c
>@@ -302,11 +302,11 @@ done:
> void ctdb_node_dead(struct ctdb_node *node)
> {
> 	if (node->ctdb->methods == NULL) {
>-		DEBUG(DEBUG_ERR,(__location__ " Can not restart transport while shutting down daemon.\n"));
>+		DBG_ERR("Can not restart transport while shutting down\n");
> 		return;
> 	}
>-
> 	node->ctdb->methods->restart(node);
>+
> 	if (node->flags & NODE_FLAGS_DISCONNECTED) {
> 		DEBUG(DEBUG_INFO,("%s: node %s is already marked disconnected: %u connected\n", 
> 			 node->ctdb->name, node->name, 
>-- 
>2.25.1
>
>
>From 710c95304dec39b37119c8462eb1e550e9b7cc8f Mon Sep 17 00:00:00 2001
>From: Noel Power <noel.power@suse.com>
>Date: Sat, 29 Feb 2020 15:49:28 +0000
>Subject: [PATCH 05/11] ctdb-tcp: move free of inbound queue to TCP restart
>
>Since commit 77deaadca8e8dbc3c92ea16893099c72f6dc874e, a nodeA which
>had previously accepted a connection from nodeB (where nodeB dies
>e.g. as as result of fencing) when nodeB attempts to connect again
>after restarting is always rejected with
>
> ctdb_listen_event: Incoming queue active, rejecting connection from w.x.y.z
>
>messages.
>
>Consolidate dead node handling in the TCP restart handling.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Noel Power <noel.power@suse.com>
>Reviewed-by: Ralph Boehme <slow@samba.org>
>Reviewed-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit 0ff1b78fc2f0491f9e11131d0040bdaba8873770)
>---
> ctdb/tcp/tcp_init.c | 2 +-
> ctdb/tcp/tcp_io.c   | 1 -
> 2 files changed, 1 insertion(+), 2 deletions(-)
>
>diff --git a/ctdb/tcp/tcp_init.c b/ctdb/tcp/tcp_init.c
>index 5bd9e2f3a09..7ea079d2303 100644
>--- a/ctdb/tcp/tcp_init.c
>+++ b/ctdb/tcp/tcp_init.c
>@@ -121,7 +121,7 @@ static void ctdb_tcp_restart(struct ctdb_node *node)
> 		node->transport_data, struct ctdb_tcp_node);
> 
> 	DEBUG(DEBUG_NOTICE,("Tearing down connection to dead node :%d\n", node->pnn));
>-
>+	TALLOC_FREE(tnode->in_queue);
> 	ctdb_tcp_stop_connection(node);
> 
> 	tnode->connect_te = tevent_add_timer(node->ctdb->ev, tnode,
>diff --git a/ctdb/tcp/tcp_io.c b/ctdb/tcp/tcp_io.c
>index df9ca02b413..bcb18fbf300 100644
>--- a/ctdb/tcp/tcp_io.c
>+++ b/ctdb/tcp/tcp_io.c
>@@ -75,7 +75,6 @@ void ctdb_tcp_read_cb(uint8_t *data, size_t cnt, void *args)
> 	return;
> 
> failed:
>-	TALLOC_FREE(tnode->in_queue);
> 	node->ctdb->upcalls->node_dead(node);
> 
> 	TALLOC_FREE(data);
>-- 
>2.25.1
>
>
>From 3f6a4db993259e1e4e2effde33c4d4002f8a0e0f Mon Sep 17 00:00:00 2001
>From: Ralph Boehme <slow@samba.org>
>Date: Sat, 29 Feb 2020 12:13:12 +0100
>Subject: [PATCH 06/11] ctdb-tcp: always call node_dead() upcall in
> ctdb_tcp_tnode_cb()
>
>ctdb_tcp_tnode_cb() is called when we receive data on the outgoing connection.
>
>This can happen when we get an EOF on the connection because the other side as
>closed. In this case data will be NULL.
>
>It would also be called if we received data from the peer. In this case data
>will not be NULL.
>
>The latter case is a fatal error though and we already call
>ctdb_tcp_stop_connection() for this case as well, which means even though the
>node is not fully connected anymore, by not calling the node_dead() upcall
>NODE_FLAGS_DISCONNECTED will not be set.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Ralph Boehme <slow@samba.org>
>Reviewed-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit b83ef98c7466b2a81968555de83fb977bb6ca9f0)
>---
> ctdb/tcp/tcp_connect.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 3e9cf062184..758d4140d43 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -65,9 +65,7 @@ void ctdb_tcp_tnode_cb(uint8_t *data, size_t cnt, void *private_data)
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
> 		node->transport_data, struct ctdb_tcp_node);
> 
>-	if (data == NULL) {
>-		node->ctdb->upcalls->node_dead(node);
>-	}
>+	node->ctdb->upcalls->node_dead(node);
> 
> 	ctdb_tcp_stop_connection(node);
> 	tnode->connect_te = tevent_add_timer(node->ctdb->ev, tnode,
>-- 
>2.25.1
>
>
>From 9d1350943261a7756af109394700293e7ac7b0ac Mon Sep 17 00:00:00 2001
>From: Ralph Boehme <slow@samba.org>
>Date: Sat, 29 Feb 2020 12:28:20 +0100
>Subject: [PATCH 07/11] ctdb-tcp: Remove redundant restart in
> ctdb_tcp_tnode_cb()
>
>The node dead upcall has already restarted the outgoing connection.
>There's no need to repeat it.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Ralph Boehme <slow@samba.org>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit ea37ecdcd5960311f54a7a5510b88a654da23daa)
>---
> ctdb/tcp/tcp_connect.c | 6 ------
> 1 file changed, 6 deletions(-)
>
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 758d4140d43..60e43db0306 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -62,15 +62,9 @@ void ctdb_tcp_stop_connection(struct ctdb_node *node)
> void ctdb_tcp_tnode_cb(uint8_t *data, size_t cnt, void *private_data)
> {
> 	struct ctdb_node *node = talloc_get_type(private_data, struct ctdb_node);
>-	struct ctdb_tcp_node *tnode = talloc_get_type(
>-		node->transport_data, struct ctdb_tcp_node);
> 
> 	node->ctdb->upcalls->node_dead(node);
> 
>-	ctdb_tcp_stop_connection(node);
>-	tnode->connect_te = tevent_add_timer(node->ctdb->ev, tnode,
>-					     timeval_current_ofs(3, 0),
>-					     ctdb_tcp_node_connect, node);
> 	TALLOC_FREE(data);
> }
> 
>-- 
>2.25.1
>
>
>From e305725e2425f68b082c63fffacc87c2d9f165b4 Mon Sep 17 00:00:00 2001
>From: Ralph Boehme <slow@samba.org>
>Date: Fri, 28 Feb 2020 11:36:00 +0100
>Subject: [PATCH 08/11] ctdb-tcp: rename ctdb_tcp_stop_connection() to
> ctdb_tcp_stop_outgoing()
>
>No change in behaviour.  This makes the code self-documenting.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Ralph Boehme <slow@samba.org>
>Reviewed-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit 1e2a967ff41cc29c3a0d7f61a46937c68fdb90ba)
>---
> ctdb/tcp/ctdb_tcp.h    |  2 +-
> ctdb/tcp/tcp_connect.c | 12 ++++++------
> ctdb/tcp/tcp_init.c    |  2 +-
> 3 files changed, 8 insertions(+), 8 deletions(-)
>
>diff --git a/ctdb/tcp/ctdb_tcp.h b/ctdb/tcp/ctdb_tcp.h
>index daabad74297..095056e8544 100644
>--- a/ctdb/tcp/ctdb_tcp.h
>+++ b/ctdb/tcp/ctdb_tcp.h
>@@ -48,7 +48,7 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
> 			   struct timeval t, void *private_data);
> void ctdb_tcp_read_cb(uint8_t *data, size_t cnt, void *args);
> void ctdb_tcp_tnode_cb(uint8_t *data, size_t cnt, void *private_data);
>-void ctdb_tcp_stop_connection(struct ctdb_node *node);
>+void ctdb_tcp_stop_outgoing(struct ctdb_node *node);
> 
> #define CTDB_TCP_ALIGNMENT 8
> 
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 60e43db0306..2a314d6554b 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -38,9 +38,9 @@
> #include "ctdb_tcp.h"
> 
> /*
>-  stop any connecting (established or pending) to a node
>+  stop any outgoing connection (established or pending) to a node
>  */
>-void ctdb_tcp_stop_connection(struct ctdb_node *node)
>+void ctdb_tcp_stop_outgoing(struct ctdb_node *node)
> {
> 	struct ctdb_tcp_node *tnode = talloc_get_type(
> 		node->transport_data, struct ctdb_tcp_node);
>@@ -90,7 +90,7 @@ static void ctdb_node_connect_write(struct tevent_context *ev,
> 
> 	ret = getsockopt(tnode->out_fd, SOL_SOCKET, SO_ERROR, &error, &len);
> 	if (ret != 0 || error != 0) {
>-		ctdb_tcp_stop_connection(node);
>+		ctdb_tcp_stop_outgoing(node);
> 		tnode->connect_te = tevent_add_timer(ctdb->ev, tnode,
> 						    timeval_current_ofs(1, 0),
> 						    ctdb_tcp_node_connect, node);
>@@ -128,7 +128,7 @@ static void ctdb_node_connect_write(struct tevent_context *ev,
> 					    node->name);
> 	if (tnode->out_queue == NULL) {
> 		DBG_ERR("Failed to set up outgoing queue\n");
>-		ctdb_tcp_stop_connection(node);
>+		ctdb_tcp_stop_outgoing(node);
> 		tnode->connect_te = tevent_add_timer(ctdb->ev,
> 						     tnode,
> 						     timeval_current_ofs(1, 0),
>@@ -168,7 +168,7 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
>         ctdb_sock_addr sock_out;
> 	int ret;
> 
>-	ctdb_tcp_stop_connection(node);
>+	ctdb_tcp_stop_outgoing(node);
> 
> 	sock_out = node->address;
> 
>@@ -252,7 +252,7 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
> 	return;
> 
> failed:
>-	ctdb_tcp_stop_connection(node);
>+	ctdb_tcp_stop_outgoing(node);
> 	tnode->connect_te = tevent_add_timer(ctdb->ev,
> 					     tnode,
> 					     timeval_current_ofs(1, 0),
>diff --git a/ctdb/tcp/tcp_init.c b/ctdb/tcp/tcp_init.c
>index 7ea079d2303..83a718172ac 100644
>--- a/ctdb/tcp/tcp_init.c
>+++ b/ctdb/tcp/tcp_init.c
>@@ -122,7 +122,7 @@ static void ctdb_tcp_restart(struct ctdb_node *node)
> 
> 	DEBUG(DEBUG_NOTICE,("Tearing down connection to dead node :%d\n", node->pnn));
> 	TALLOC_FREE(tnode->in_queue);
>-	ctdb_tcp_stop_connection(node);
>+	ctdb_tcp_stop_outgoing(node);
> 
> 	tnode->connect_te = tevent_add_timer(node->ctdb->ev, tnode,
> 					     timeval_zero(),
>-- 
>2.25.1
>
>
>From fbd3320f21d18eacb8c842a078d40148d154e4ed Mon Sep 17 00:00:00 2001
>From: Ralph Boehme <slow@samba.org>
>Date: Sat, 29 Feb 2020 11:54:51 +0100
>Subject: [PATCH 09/11] ctdb-tcp: add ctdb_tcp_stop_incoming()
>
>No change in behaviour.  This makes the code self-documenting.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Ralph Boehme <slow@samba.org>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit 2c73dbafba50b28e72a8ec7b4382fae42fca6d17)
>---
> ctdb/tcp/ctdb_tcp.h    |  1 +
> ctdb/tcp/tcp_connect.c | 10 ++++++++++
> ctdb/tcp/tcp_init.c    |  2 +-
> 3 files changed, 12 insertions(+), 1 deletion(-)
>
>diff --git a/ctdb/tcp/ctdb_tcp.h b/ctdb/tcp/ctdb_tcp.h
>index 095056e8544..cb8d66fa5dc 100644
>--- a/ctdb/tcp/ctdb_tcp.h
>+++ b/ctdb/tcp/ctdb_tcp.h
>@@ -49,6 +49,7 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
> void ctdb_tcp_read_cb(uint8_t *data, size_t cnt, void *args);
> void ctdb_tcp_tnode_cb(uint8_t *data, size_t cnt, void *private_data);
> void ctdb_tcp_stop_outgoing(struct ctdb_node *node);
>+void ctdb_tcp_stop_incoming(struct ctdb_node *node);
> 
> #define CTDB_TCP_ALIGNMENT 8
> 
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index 2a314d6554b..e8ac148baa1 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -54,6 +54,16 @@ void ctdb_tcp_stop_outgoing(struct ctdb_node *node)
> 	}
> }
> 
>+/*
>+  stop incoming connection to a node
>+ */
>+void ctdb_tcp_stop_incoming(struct ctdb_node *node)
>+{
>+	struct ctdb_tcp_node *tnode = talloc_get_type(
>+		node->transport_data, struct ctdb_tcp_node);
>+
>+	TALLOC_FREE(tnode->in_queue);
>+}
> 
> /*
>   called when a complete packet has come in - should not happen on this socket
>diff --git a/ctdb/tcp/tcp_init.c b/ctdb/tcp/tcp_init.c
>index 83a718172ac..3c99c8b31f5 100644
>--- a/ctdb/tcp/tcp_init.c
>+++ b/ctdb/tcp/tcp_init.c
>@@ -121,7 +121,7 @@ static void ctdb_tcp_restart(struct ctdb_node *node)
> 		node->transport_data, struct ctdb_tcp_node);
> 
> 	DEBUG(DEBUG_NOTICE,("Tearing down connection to dead node :%d\n", node->pnn));
>-	TALLOC_FREE(tnode->in_queue);
>+	ctdb_tcp_stop_incoming(node);
> 	ctdb_tcp_stop_outgoing(node);
> 
> 	tnode->connect_te = tevent_add_timer(node->ctdb->ev, tnode,
>-- 
>2.25.1
>
>
>From 1af80bb722457b454772e0e86a23897480f38230 Mon Sep 17 00:00:00 2001
>From: Martin Schwenke <martin@meltin.net>
>Date: Fri, 6 Mar 2020 15:59:32 +1100
>Subject: [PATCH 10/11] ctdb-tcp: Factor out function ctdb_tcp_start_outgoing()
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Amitay Isaacs <amitay@gmail.com>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>(cherry picked from commit 3c8747fe29486a4f95308b335a5e3ec1807f62cb)
>---
> ctdb/tcp/tcp_connect.c | 19 +++++++++++++------
> 1 file changed, 13 insertions(+), 6 deletions(-)
>
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index e8ac148baa1..a7bda43ebbf 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -164,11 +164,8 @@ static void ctdb_node_connect_write(struct tevent_context *ev,
> /*
>   called when we should try and establish a tcp connection to a node
> */
>-void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
>-			   struct timeval t, void *private_data)
>+static void ctdb_tcp_start_outgoing(struct ctdb_node *node)
> {
>-	struct ctdb_node *node = talloc_get_type(private_data,
>-						 struct ctdb_node);
> 	struct ctdb_tcp_node *tnode = talloc_get_type(node->transport_data,
> 						      struct ctdb_tcp_node);
> 	struct ctdb_context *ctdb = node->ctdb;
>@@ -178,8 +175,6 @@ void ctdb_tcp_node_connect(struct tevent_context *ev, struct tevent_timer *te,
>         ctdb_sock_addr sock_out;
> 	int ret;
> 
>-	ctdb_tcp_stop_outgoing(node);
>-
> 	sock_out = node->address;
> 
> 	tnode->out_fd = socket(sock_out.sa.sa_family, SOCK_STREAM, IPPROTO_TCP);
>@@ -270,6 +265,18 @@ failed:
> 					     node);
> }
> 
>+void ctdb_tcp_node_connect(struct tevent_context *ev,
>+			   struct tevent_timer *te,
>+			   struct timeval t,
>+			   void *private_data)
>+{
>+	struct ctdb_node *node = talloc_get_type_abort(private_data,
>+						       struct ctdb_node);
>+
>+	ctdb_tcp_stop_outgoing(node);
>+	ctdb_tcp_start_outgoing(node);
>+}
>+
> /*
>   called when we get contacted by another node
>   currently makes no attempt to check if the connection is really from a ctdb
>-- 
>2.25.1
>
>
>From 783335f207d73ebe81ff514452c535b2906ea3c3 Mon Sep 17 00:00:00 2001
>From: Martin Schwenke <martin@meltin.net>
>Date: Fri, 6 Mar 2020 16:11:23 +1100
>Subject: [PATCH 11/11] ctdb-tcp: Do not stop outbound connection in
> ctdb_tcp_node_connect()
>
>The only place the outgoing connection needs to be stopped is when
>there is a timeout when waiting for the connection to become writable.
>Add a new function ctdb_tcp_node_connect_timeout() to handle this
>case.
>
>All of the other cases are attempts to establish a new outgoing
>connection (initial attempt, retry after an error or disconnect, ...)
>so drop stopping the connection in those cases.
>
>BUG: https://bugzilla.samba.org/show_bug.cgi?id=14295
>
>Signed-off-by: Amitay Isaacs <amitay@gmail.com>
>Signed-off-by: Martin Schwenke <martin@meltin.net>
>
>Autobuild-User(master): Martin Schwenke <martins@samba.org>
>Autobuild-Date(master): Thu Mar 12 05:29:20 UTC 2020 on sn-devel-184
>
>(cherry picked from commit 319c93f0c6a949545229b616dfbd4f51baf11171)
>---
> ctdb/tcp/tcp_connect.c | 18 +++++++++++++++++-
> 1 file changed, 17 insertions(+), 1 deletion(-)
>
>diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
>index a7bda43ebbf..d9bc52aa22f 100644
>--- a/ctdb/tcp/tcp_connect.c
>+++ b/ctdb/tcp/tcp_connect.c
>@@ -161,6 +161,11 @@ static void ctdb_node_connect_write(struct tevent_context *ev,
> }
> 
> 
>+static void ctdb_tcp_node_connect_timeout(struct tevent_context *ev,
>+					  struct tevent_timer *te,
>+					  struct timeval t,
>+					  void *private_data);
>+
> /*
>   called when we should try and establish a tcp connection to a node
> */
>@@ -251,7 +256,7 @@ static void ctdb_tcp_start_outgoing(struct ctdb_node *node)
> 	tnode->connect_te = tevent_add_timer(ctdb->ev,
> 					     tnode,
> 					     timeval_current_ofs(1, 0),
>-					     ctdb_tcp_node_connect,
>+					     ctdb_tcp_node_connect_timeout,
> 					     node);
> 
> 	return;
>@@ -273,6 +278,17 @@ void ctdb_tcp_node_connect(struct tevent_context *ev,
> 	struct ctdb_node *node = talloc_get_type_abort(private_data,
> 						       struct ctdb_node);
> 
>+	ctdb_tcp_start_outgoing(node);
>+}
>+
>+static void ctdb_tcp_node_connect_timeout(struct tevent_context *ev,
>+					  struct tevent_timer *te,
>+					  struct timeval t,
>+					  void *private_data)
>+{
>+	struct ctdb_node *node = talloc_get_type_abort(private_data,
>+						       struct ctdb_node);
>+
> 	ctdb_tcp_stop_outgoing(node);
> 	ctdb_tcp_start_outgoing(node);
> }
>-- 
>2.25.1
>

Flags:

slow: review+

Actions: View

Attachments on bug 14295: 15823 | 15826 | 15827 | 15828 | 15829 | 15830 | 15855 | 15859 | 15860 | 15862 | 15869