[sheepdog] [PATCH 1/2] sheep: set keepalive in all client socket fd

[Yunkai Zhang]

From: Yunkai Zhang <qiushu.zyk at taobao.com>

We have set keepalive in cached socket fd, but uncached fd should also
need to set this option. So I move set_keepalive() from get_sheep_fd()
to connect_to().

Signed-off-by: Yunkai Zhang <qiushu.zyk at taobao.com>
---
 include/net.h |    1 +
 lib/net.c     |   53 +++++++++++++++++++++++++++++++++++++++++++++++++----
 sheep/sdnet.c |   47 +----------------------------------------------
 3 files changed, 51 insertions(+), 50 deletions(-)

diff --git a/include/net.h b/include/net.h
index 83da12a..581da66 100644
--- a/include/net.h
+++ b/include/net.h
@@ -52,6 +52,7 @@ char *addr_to_str(char *str, int size, uint8_t *addr, uint16_t port);
 uint8_t *str_to_addr(int af, const char *ipstr, uint8_t *addr);
 int set_nonblocking(int fd);
 int set_nodelay(int fd);
+int set_keepalive(int fd);
 int set_timeout(int fd);
 int get_local_addr(uint8_t *bytes);
 
diff --git a/lib/net.c b/lib/net.c
index c4a96ac..a7deb2a 100644
--- a/lib/net.c
+++ b/lib/net.c
@@ -229,13 +229,26 @@ int connect_to(const char *name, int port)
 		}
 
 		ret = connect(fd, res->ai_addr, res->ai_addrlen);
-		if (ret)
+		if (ret) {
 			eprintf("failed to connect to %s:%d: %m\n",
 				name, port);
-		else
-			goto success;
+			close(fd);
+			continue;
+		}
 
-		close(fd);
+		ret = set_keepalive(fd);
+		if (ret) {
+			close(fd);
+			break;
+		}
+
+		ret = set_nodelay(fd);
+		if (ret) {
+			eprintf("%m\n");
+			close(fd);
+			break;
+		} else
+			goto success;
 	}
 	fd = -1;
 success:
@@ -420,6 +433,38 @@ int set_nodelay(int fd)
 	return ret;
 }
 
+/*
+ * Timeout after request is issued after 5s.
+ *
+ * Heart-beat message will be sent periodically with 1s interval.
+ * If the node of the other end of fd fails, we'll detect it in 3s
+ */
+int set_keepalive(int fd)
+{
+	int val = 1;
+
+	if (setsockopt(fd, SOL_SOCKET, SO_KEEPALIVE, &val, sizeof(val)) < 0) {
+		dprintf("%m\n");
+		return -1;
+	}
+	val = 5;
+	if (setsockopt(fd, SOL_TCP, TCP_KEEPIDLE, &val, sizeof(val)) < 0) {
+		dprintf("%m\n");
+		return -1;
+	}
+	val = 1;
+	if (setsockopt(fd, SOL_TCP, TCP_KEEPINTVL, &val, sizeof(val)) < 0) {
+		dprintf("%m\n");
+		return -1;
+	}
+	val = 3;
+	if (setsockopt(fd, SOL_TCP, TCP_KEEPCNT, &val, sizeof(val)) < 0) {
+		dprintf("%m\n");
+		return -1;
+	}
+	return 0;
+}
+
 int get_local_addr(uint8_t *bytes)
 {
 	struct ifaddrs *ifaddr, *ifa;
diff --git a/sheep/sdnet.c b/sheep/sdnet.c
index 8479a74..3ae260e 100644
--- a/sheep/sdnet.c
+++ b/sheep/sdnet.c
@@ -828,41 +828,9 @@ void del_sheep_fd(int fd)
 	}
 }
 
-/*
- * Timeout after request is issued after 5s.
- *
- * Heart-beat message will be sent periodically with 1s interval.
- * If the node of the other end of fd fails, we'll detect it in 3s
- */
-static int set_keepalive(int fd)
-{
-	int val = 1;
-
-	if (setsockopt(fd, SOL_SOCKET, SO_KEEPALIVE, &val, sizeof(val)) < 0) {
-		dprintf("%m\n");
-		return -1;
-	}
-	val = 5;
-	if (setsockopt(fd, SOL_TCP, TCP_KEEPIDLE, &val, sizeof(val)) < 0) {
-		dprintf("%m\n");
-		return -1;
-	}
-	val = 1;
-	if (setsockopt(fd, SOL_TCP, TCP_KEEPINTVL, &val, sizeof(val)) < 0) {
-		dprintf("%m\n");
-		return -1;
-	}
-	val = 3;
-	if (setsockopt(fd, SOL_TCP, TCP_KEEPCNT, &val, sizeof(val)) < 0) {
-		dprintf("%m\n");
-		return -1;
-	}
-	return 0;
-}
-
 int get_sheep_fd(uint8_t *addr, uint16_t port, int node_idx, uint32_t epoch)
 {
-	int i, fd, ret;
+	int i, fd;
 	char name[INET6_ADDRSTRLEN];
 
 	if (cached_epoch == 0) {
@@ -902,19 +870,6 @@ int get_sheep_fd(uint8_t *addr, uint16_t port, int node_idx, uint32_t epoch)
 	if (fd < 0)
 		return -1;
 
-	ret = set_keepalive(fd);
-	if (ret) {
-		close(fd);
-		return -1;
-	}
-
-	ret = set_nodelay(fd);
-	if (ret) {
-		eprintf("%m\n");
-		close(fd);
-		return -1;
-	}
-
 	cached_fds[node_idx] = fd;
 
 	return fd;
-- 
1.7.10.2