From 53706b41b930fe6374add3ad097a627254c45bc1 Mon Sep 17 00:00:00 2001
From: David Bigagli <david@schedmd.com>
Date: Fri, 27 Jun 2014 11:43:35 -0700
Subject: [PATCH] Ignore transient cgroup errors.
---
NEWS | 2 ++
src/plugins/proctrack/cgroup/proctrack_cgroup.c | 5 ++++-
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/NEWS b/NEWS
index 99481b0e643..0fa36a005b6 100644
--- a/NEWS
+++ b/NEWS
@@ -29,6 +29,8 @@ documents those changes that are of interest to users and admins.
you could get multiple '*' wckeys.
-- Fix bug which could report to squeue the wrong partition for a running job
that is submitted to multiple partitions.
+ -- If failed to rmdir() a cgroup ignore the error preventing slurmstepd to
+ loop forever.
* Changes in Slurm 14.03.4
==========================
diff --git a/src/plugins/proctrack/cgroup/proctrack_cgroup.c b/src/plugins/proctrack/cgroup/proctrack_cgroup.c
index d3608d3dd93..e92386937d6 100644
--- a/src/plugins/proctrack/cgroup/proctrack_cgroup.c
+++ b/src/plugins/proctrack/cgroup/proctrack_cgroup.c
@@ -287,8 +287,11 @@ int _slurm_cgroup_destroy(void)
if (xcgroup_delete(&step_freezer_cg) != XCGROUP_SUCCESS) {
error("_slurm_cgroup_destroy: problem deleting step "
"cgroup path %s: %m", step_freezer_cg.path);
+ /* Ignore the error since the cgroup will
+ * be eventually cleaned up by the release
+ * script.
+ */
xcgroup_unlock(&freezer_cg);
- return SLURM_ERROR;
}
xcgroup_destroy(&step_freezer_cg);
}
--
GitLab