svn merge -r16594:16610 https://eris.llnl.gov/svn/slurm/branches/slurm-1.3

NEWS

@@ -181,6 +181,8 @@ documents those changes that are of interest to users and admins.
     STARTJOB request: "TASKLIST includes non-responsive nodes".
  -- Fix bug in select/linear when used with sched/gang that can result in a 
     job's required or excluded node specification being ignored.
+ -- Add logic to handle message connect timeouts (timed-out.patch from 
+    Chuck Clouston, Bull).
 
 * Changes in SLURM 1.3.13
 =========================

src/common/slurm_protocol_socket_implementation.c

@@ -77,7 +77,7 @@
 #include "src/common/xmalloc.h"
 #include "src/common/util-net.h"
 
-#define PORT_RETRIES    2
+#define PORT_RETRIES    3
 #define MIN_USER_PORT   (IPPORT_RESERVED + 1)
 #define MAX_USER_PORT   0xffff
 #define RANDOM_USER_PORT ((uint16_t) ((lrand48() % \
@@ -525,7 +525,7 @@ slurm_fd _slurm_open_stream(slurm_addr *addr, bool retry)
 				    sizeof(*addr));
 		if (rc >= 0)		    /* success */
 			break;
-		if ((errno != ECONNREFUSED) || 
+		if (((errno != ECONNREFUSED) && (errno != ETIMEDOUT)) ||
 		    (!retry) || (retry_cnt >= PORT_RETRIES)) {
 			slurm_seterrno(errno);
 			goto error;