From 56c47518d25bafc98f272b39f654020f512fd582 Mon Sep 17 00:00:00 2001
From: Moe Jette <jette1@llnl.gov>
Date: Wed, 21 Mar 2007 18:02:03 +0000
Subject: [PATCH] Add special message for user root to restart slurmctld
daemons as needed. I keep finding the backup slurmctld down on production
clusters and hope this helps.
---
src/scontrol/scontrol.c | 19 +++++++++++++++----
1 file changed, 15 insertions(+), 4 deletions(-)
diff --git a/src/scontrol/scontrol.c b/src/scontrol/scontrol.c
index 990be525474..4727a1867e3 100644
--- a/src/scontrol/scontrol.c
+++ b/src/scontrol/scontrol.c
@@ -339,6 +339,7 @@ _ping_slurmctld(char *control_machine, char *backup_controller)
{
static char *state[2] = { "UP", "DOWN" };
int primary = 1, secondary = 1;
+ int down_msg = 0;
if (slurm_ping(1) == SLURM_SUCCESS)
primary = 0;
@@ -347,17 +348,27 @@ _ping_slurmctld(char *control_machine, char *backup_controller)
fprintf(stdout, "Slurmctld(primary/backup) ");
if (control_machine || backup_controller) {
fprintf(stdout, "at ");
- if (control_machine)
+ if (control_machine) {
fprintf(stdout, "%s/", control_machine);
- else
+ if (primary)
+ down_msg = 1;
+ } else
fprintf(stdout, "(NULL)/");
- if (backup_controller)
+ if (backup_controller) {
fprintf(stdout, "%s ", backup_controller);
- else
+ if (secondary)
+ down_msg = 1;
+ } else
fprintf(stdout, "(NULL) ");
}
fprintf(stdout, "are %s/%s\n",
state[primary], state[secondary]);
+
+ if (down_msg && ((getuid() == 0) || (geteuid() == 0))) {
+ fprintf(stdout, "*****************************************\n");
+ fprintf(stdout, "** RESTORE SLURMCTLD DAEMON TO SERVICE **\n");
+ fprintf(stdout, "*****************************************\n");
+ }
}
/*
--
GitLab