summaryrefslogtreecommitdiffstats
path: root/ctdb/tcp/tcp_connect.c
diff options
context:
space:
mode:
Diffstat (limited to 'ctdb/tcp/tcp_connect.c')
-rw-r--r--ctdb/tcp/tcp_connect.c599
1 files changed, 599 insertions, 0 deletions
diff --git a/ctdb/tcp/tcp_connect.c b/ctdb/tcp/tcp_connect.c
new file mode 100644
index 0000000..6f5862a
--- /dev/null
+++ b/ctdb/tcp/tcp_connect.c
@@ -0,0 +1,599 @@
+/*
+ ctdb over TCP
+
+ Copyright (C) Andrew Tridgell 2006
+ Copyright (C) Ronnie Sahlberg 2008
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, see <http://www.gnu.org/licenses/>.
+*/
+
+#include "replace.h"
+#include "system/network.h"
+#include "system/filesys.h"
+
+#include <talloc.h>
+#include <tevent.h>
+
+#include "lib/util/debug.h"
+#include "lib/util/time.h"
+#include "lib/util/blocking.h"
+
+#include "ctdb_private.h"
+
+#include "common/system.h"
+#include "common/common.h"
+#include "common/logging.h"
+
+#include "ctdb_tcp.h"
+
+/*
+ stop any outgoing connection (established or pending) to a 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);
+
+ TALLOC_FREE(tnode->out_queue);
+ TALLOC_FREE(tnode->connect_te);
+ TALLOC_FREE(tnode->connect_fde);
+ if (tnode->out_fd != -1) {
+ close(tnode->out_fd);
+ tnode->out_fd = -1;
+ }
+}
+
+/*
+ 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
+ unless the other side closes the connection with RST or FIN
+ */
+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);
+
+ node->ctdb->upcalls->node_dead(node);
+
+ TALLOC_FREE(data);
+}
+
+/*
+ called when socket becomes writeable on connect
+*/
+static void ctdb_node_connect_write(struct tevent_context *ev,
+ struct tevent_fd *fde,
+ uint16_t flags, 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);
+ struct ctdb_context *ctdb = node->ctdb;
+ int error = 0;
+ socklen_t len = sizeof(error);
+ int one = 1;
+ int ret;
+
+ TALLOC_FREE(tnode->connect_te);
+
+ ret = getsockopt(tnode->out_fd, SOL_SOCKET, SO_ERROR, &error, &len);
+ if (ret != 0 || error != 0) {
+ ctdb_tcp_stop_outgoing(node);
+ tnode->connect_te = tevent_add_timer(ctdb->ev, tnode,
+ timeval_current_ofs(1, 0),
+ ctdb_tcp_node_connect, node);
+ return;
+ }
+
+ TALLOC_FREE(tnode->connect_fde);
+
+ ret = setsockopt(tnode->out_fd,
+ IPPROTO_TCP,
+ TCP_NODELAY,
+ (char *)&one,
+ sizeof(one));
+ if (ret == -1) {
+ DBG_WARNING("Failed to set TCP_NODELAY on fd - %s\n",
+ strerror(errno));
+ }
+ ret = setsockopt(tnode->out_fd,
+ SOL_SOCKET,
+ SO_KEEPALIVE,(char *)&one,
+ sizeof(one));
+ if (ret == -1) {
+ DBG_WARNING("Failed to set KEEPALIVE on fd - %s\n",
+ strerror(errno));
+ }
+
+ tnode->out_queue = ctdb_queue_setup(node->ctdb,
+ tnode,
+ tnode->out_fd,
+ CTDB_TCP_ALIGNMENT,
+ ctdb_tcp_tnode_cb,
+ node,
+ "to-node-%s",
+ node->name);
+ if (tnode->out_queue == NULL) {
+ DBG_ERR("Failed to set up outgoing queue\n");
+ ctdb_tcp_stop_outgoing(node);
+ tnode->connect_te = tevent_add_timer(ctdb->ev,
+ tnode,
+ timeval_current_ofs(1, 0),
+ ctdb_tcp_node_connect,
+ node);
+ return;
+ }
+
+ /* the queue subsystem now owns this fd */
+ tnode->out_fd = -1;
+
+ /*
+ * Mark the node to which this connection has been established
+ * as connected, but only if the corresponding listening
+ * socket is also connected
+ */
+ if (tnode->in_queue != NULL) {
+ node->ctdb->upcalls->node_connected(node);
+ }
+}
+
+
+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
+*/
+static void ctdb_tcp_start_outgoing(struct ctdb_node *node)
+{
+ 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;
+ int sockin_size;
+ int sockout_size;
+ ctdb_sock_addr sock_out;
+ int ret;
+
+ sock_out = node->address;
+
+ tnode->out_fd = socket(sock_out.sa.sa_family, SOCK_STREAM, IPPROTO_TCP);
+ if (tnode->out_fd == -1) {
+ DBG_ERR("Failed to create socket\n");
+ goto failed;
+ }
+
+ ret = set_blocking(tnode->out_fd, false);
+ if (ret != 0) {
+ DBG_ERR("Failed to set socket non-blocking (%s)\n",
+ strerror(errno));
+ goto failed;
+ }
+
+ set_close_on_exec(tnode->out_fd);
+
+ DBG_DEBUG("Created TCP SOCKET FD:%d\n", tnode->out_fd);
+
+ /* Bind our side of the socketpair to the same address we use to listen
+ * on incoming CTDB traffic.
+ * We must specify this address to make sure that the address we expose to
+ * the remote side is actually routable in case CTDB traffic will run on
+ * a dedicated non-routeable network.
+ */
+ sock_in = *ctdb->address;
+
+ /* AIX libs check to see if the socket address and length
+ arguments are consistent with each other on calls like
+ connect(). Can not get by with just sizeof(sock_in),
+ need sizeof(sock_in.ip).
+ */
+ switch (sock_in.sa.sa_family) {
+ case AF_INET:
+ sock_in.ip.sin_port = 0 /* Any port */;
+ sockin_size = sizeof(sock_in.ip);
+ sockout_size = sizeof(sock_out.ip);
+ break;
+ case AF_INET6:
+ sock_in.ip6.sin6_port = 0 /* Any port */;
+ sockin_size = sizeof(sock_in.ip6);
+ sockout_size = sizeof(sock_out.ip6);
+ break;
+ default:
+ DBG_ERR("Unknown address family %u\n", sock_in.sa.sa_family);
+ /* Can't happen to due to address parsing restrictions */
+ goto failed;
+ }
+
+ ret = bind(tnode->out_fd, (struct sockaddr *)&sock_in, sockin_size);
+ if (ret == -1) {
+ DBG_ERR("Failed to bind socket (%s)\n", strerror(errno));
+ goto failed;
+ }
+
+ ret = connect(tnode->out_fd,
+ (struct sockaddr *)&sock_out,
+ sockout_size);
+ if (ret != 0 && errno != EINPROGRESS) {
+ goto failed;
+ }
+
+ /* non-blocking connect - wait for write event */
+ tnode->connect_fde = tevent_add_fd(node->ctdb->ev,
+ tnode,
+ tnode->out_fd,
+ TEVENT_FD_WRITE|TEVENT_FD_READ,
+ ctdb_node_connect_write,
+ node);
+
+ /* don't give it long to connect - retry in one second. This ensures
+ that we find a node is up quickly (tcp normally backs off a syn reply
+ delay by quite a lot) */
+ tnode->connect_te = tevent_add_timer(ctdb->ev,
+ tnode,
+ timeval_current_ofs(1, 0),
+ ctdb_tcp_node_connect_timeout,
+ node);
+
+ return;
+
+failed:
+ ctdb_tcp_stop_outgoing(node);
+ tnode->connect_te = tevent_add_timer(ctdb->ev,
+ tnode,
+ timeval_current_ofs(1, 0),
+ ctdb_tcp_node_connect,
+ 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_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);
+}
+
+/*
+ called when we get contacted by another node
+ currently makes no attempt to check if the connection is really from a ctdb
+ node in our cluster
+*/
+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->transport_data,
+ struct ctdb_tcp);
+ ctdb_sock_addr addr;
+ socklen_t len;
+ int fd;
+ struct ctdb_node *node;
+ struct ctdb_tcp_node *tnode;
+ int one = 1;
+ int ret;
+
+ memset(&addr, 0, sizeof(addr));
+ len = sizeof(addr);
+ fd = accept(ctcp->listen_fd, (struct sockaddr *)&addr, &len);
+ if (fd == -1) return;
+ smb_set_close_on_exec(fd);
+
+ node = ctdb_ip_to_node(ctdb, &addr);
+ if (node == NULL) {
+ D_ERR("Refused connection from unknown node %s\n",
+ ctdb_addr_to_str(&addr));
+ close(fd);
+ return;
+ }
+
+ tnode = talloc_get_type_abort(node->transport_data,
+ struct ctdb_tcp_node);
+ if (tnode == NULL) {
+ /* This can't happen - see ctdb_tcp_initialise() */
+ DBG_ERR("INTERNAL ERROR setting up connection from node %s\n",
+ ctdb_addr_to_str(&addr));
+ close(fd);
+ return;
+ }
+
+ if (tnode->in_queue != NULL) {
+ DBG_ERR("Incoming queue active, rejecting connection from %s\n",
+ ctdb_addr_to_str(&addr));
+ close(fd);
+ return;
+ }
+
+ ret = set_blocking(fd, false);
+ if (ret != 0) {
+ DBG_ERR("Failed to set socket non-blocking (%s)\n",
+ strerror(errno));
+ close(fd);
+ return;
+ }
+
+ set_close_on_exec(fd);
+
+ DBG_DEBUG("Created SOCKET FD:%d to incoming ctdb connection\n", fd);
+
+ ret = setsockopt(fd,
+ SOL_SOCKET,
+ SO_KEEPALIVE,
+ (char *)&one,
+ sizeof(one));
+ if (ret == -1) {
+ DBG_WARNING("Failed to set KEEPALIVE on fd - %s\n",
+ strerror(errno));
+ }
+
+ tnode->in_queue = ctdb_queue_setup(ctdb,
+ tnode,
+ fd,
+ CTDB_TCP_ALIGNMENT,
+ ctdb_tcp_read_cb,
+ node,
+ "ctdbd-%s",
+ ctdb_addr_to_str(&addr));
+ if (tnode->in_queue == NULL) {
+ DBG_ERR("Failed to set up incoming queue\n");
+ close(fd);
+ return;
+ }
+
+ /*
+ * Mark the connecting node as connected, but only if the
+ * corresponding outbound connected is also up
+ */
+ if (tnode->out_queue != NULL) {
+ node->ctdb->upcalls->node_connected(node);
+ }
+ }
+
+
+/*
+ automatically find which address to listen on
+*/
+static int ctdb_tcp_listen_automatic(struct ctdb_context *ctdb)
+{
+ struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
+ struct ctdb_tcp);
+ ctdb_sock_addr sock;
+ int lock_fd;
+ unsigned int i;
+ const char *lock_path = CTDB_RUNDIR "/.socket_lock";
+ struct flock lock;
+ int one = 1;
+ int sock_size;
+ struct tevent_fd *fde;
+
+ /* If there are no nodes, then it won't be possible to find
+ * the first one. Log a failure and short circuit the whole
+ * process.
+ */
+ if (ctdb->num_nodes == 0) {
+ DEBUG(DEBUG_CRIT,("No nodes available to attempt bind to - is the nodes file empty?\n"));
+ return -1;
+ }
+
+ /* in order to ensure that we don't get two nodes with the
+ same address, we must make the bind() and listen() calls
+ atomic. The SO_REUSEADDR setsockopt only prevents double
+ binds if the first socket is in LISTEN state */
+ lock_fd = open(lock_path, O_RDWR|O_CREAT, 0666);
+ if (lock_fd == -1) {
+ DEBUG(DEBUG_CRIT,("Unable to open %s\n", lock_path));
+ return -1;
+ }
+
+ lock.l_type = F_WRLCK;
+ lock.l_whence = SEEK_SET;
+ lock.l_start = 0;
+ lock.l_len = 1;
+ lock.l_pid = 0;
+
+ if (fcntl(lock_fd, F_SETLKW, &lock) != 0) {
+ DEBUG(DEBUG_CRIT,("Unable to lock %s\n", lock_path));
+ close(lock_fd);
+ return -1;
+ }
+
+ for (i=0; i < ctdb->num_nodes; i++) {
+ if (ctdb->nodes[i]->flags & NODE_FLAGS_DELETED) {
+ continue;
+ }
+ sock = ctdb->nodes[i]->address;
+
+ switch (sock.sa.sa_family) {
+ case AF_INET:
+ sock_size = sizeof(sock.ip);
+ break;
+ case AF_INET6:
+ sock_size = sizeof(sock.ip6);
+ break;
+ default:
+ DEBUG(DEBUG_ERR, (__location__ " unknown family %u\n",
+ sock.sa.sa_family));
+ continue;
+ }
+
+ ctcp->listen_fd = socket(sock.sa.sa_family, SOCK_STREAM, IPPROTO_TCP);
+ if (ctcp->listen_fd == -1) {
+ ctdb_set_error(ctdb, "socket failed\n");
+ continue;
+ }
+
+ set_close_on_exec(ctcp->listen_fd);
+
+ if (setsockopt(ctcp->listen_fd,SOL_SOCKET,SO_REUSEADDR,
+ (char *)&one,sizeof(one)) == -1) {
+ DEBUG(DEBUG_WARNING, ("Failed to set REUSEADDR on fd - %s\n",
+ strerror(errno)));
+ }
+
+ if (bind(ctcp->listen_fd, (struct sockaddr * )&sock, sock_size) == 0) {
+ break;
+ }
+
+ if (errno == EADDRNOTAVAIL) {
+ DEBUG(DEBUG_DEBUG,(__location__ " Failed to bind() to socket. %s(%d)\n",
+ strerror(errno), errno));
+ } else {
+ DEBUG(DEBUG_ERR,(__location__ " Failed to bind() to socket. %s(%d)\n",
+ strerror(errno), errno));
+ }
+
+ close(ctcp->listen_fd);
+ ctcp->listen_fd = -1;
+ }
+
+ if (i == ctdb->num_nodes) {
+ DEBUG(DEBUG_CRIT,("Unable to bind to any of the node addresses - giving up\n"));
+ goto failed;
+ }
+ ctdb->address = talloc_memdup(ctdb,
+ &ctdb->nodes[i]->address,
+ sizeof(ctdb_sock_addr));
+ if (ctdb->address == NULL) {
+ ctdb_set_error(ctdb, "Out of memory at %s:%d",
+ __FILE__, __LINE__);
+ goto failed;
+ }
+
+ ctdb->name = talloc_asprintf(ctdb, "%s:%u",
+ ctdb_addr_to_str(ctdb->address),
+ ctdb_addr_to_port(ctdb->address));
+ if (ctdb->name == NULL) {
+ ctdb_set_error(ctdb, "Out of memory at %s:%d",
+ __FILE__, __LINE__);
+ goto failed;
+ }
+ DEBUG(DEBUG_INFO,("ctdb chose network address %s\n", ctdb->name));
+
+ if (listen(ctcp->listen_fd, 10) == -1) {
+ goto failed;
+ }
+
+ fde = tevent_add_fd(ctdb->ev, ctcp, ctcp->listen_fd, TEVENT_FD_READ,
+ ctdb_listen_event, ctdb);
+ tevent_fd_set_auto_close(fde);
+
+ close(lock_fd);
+
+ return 0;
+
+failed:
+ close(lock_fd);
+ if (ctcp->listen_fd != -1) {
+ close(ctcp->listen_fd);
+ ctcp->listen_fd = -1;
+ }
+ return -1;
+}
+
+
+/*
+ listen on our own address
+*/
+int ctdb_tcp_listen(struct ctdb_context *ctdb)
+{
+ struct ctdb_tcp *ctcp = talloc_get_type(ctdb->transport_data,
+ struct ctdb_tcp);
+ ctdb_sock_addr sock;
+ int sock_size;
+ int one = 1;
+ struct tevent_fd *fde;
+
+ /* we can either auto-bind to the first available address, or we can
+ use a specified address */
+ if (!ctdb->address) {
+ return ctdb_tcp_listen_automatic(ctdb);
+ }
+
+ sock = *ctdb->address;
+
+ switch (sock.sa.sa_family) {
+ case AF_INET:
+ sock_size = sizeof(sock.ip);
+ break;
+ case AF_INET6:
+ sock_size = sizeof(sock.ip6);
+ break;
+ default:
+ DEBUG(DEBUG_ERR, (__location__ " unknown family %u\n",
+ sock.sa.sa_family));
+ goto failed;
+ }
+
+ ctcp->listen_fd = socket(sock.sa.sa_family, SOCK_STREAM, IPPROTO_TCP);
+ if (ctcp->listen_fd == -1) {
+ ctdb_set_error(ctdb, "socket failed\n");
+ return -1;
+ }
+
+ set_close_on_exec(ctcp->listen_fd);
+
+ if (setsockopt(ctcp->listen_fd,SOL_SOCKET,SO_REUSEADDR,(char *)&one,sizeof(one)) == -1) {
+ DEBUG(DEBUG_WARNING, ("Failed to set REUSEADDR on fd - %s\n",
+ strerror(errno)));
+ }
+
+ if (bind(ctcp->listen_fd, (struct sockaddr * )&sock, sock_size) != 0) {
+ DEBUG(DEBUG_ERR,(__location__ " Failed to bind() to socket. %s(%d)\n", strerror(errno), errno));
+ goto failed;
+ }
+
+ if (listen(ctcp->listen_fd, 10) == -1) {
+ goto failed;
+ }
+
+ fde = tevent_add_fd(ctdb->ev, ctcp, ctcp->listen_fd, TEVENT_FD_READ,
+ ctdb_listen_event, ctdb);
+ tevent_fd_set_auto_close(fde);
+
+ return 0;
+
+failed:
+ if (ctcp->listen_fd != -1) {
+ close(ctcp->listen_fd);
+ }
+ ctcp->listen_fd = -1;
+ return -1;
+}
+
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-18 04:13:24 +0000