Merge 4.0.6 master->release/4.0

README.md

@@ -5,7 +5,7 @@ CycleCloud Slurm Clusters in Azure
 This project sets up an auto-scaling Slurm cluster
 Slurm is a highly configurable open source workload manager. See the [Slurm project site](https://www.schedmd.com/) for an overview.
 # Table of Contents:
-1. [Managing Slurm Clusters in 4.0.5](#managing-slurm-clusters)
+1. [Managing Slurm Clusters in 4.0.6](#managing-slurm-clusters)
     1. [Making Cluster Changes](#making-cluster-changes)
     2. [No longer pre-creating execute nodes](#no-longer-pre-creating-execute-nodes)
     3. [Creating additional partitions](#creating-additional-partitions)
@@ -36,7 +36,7 @@ Slurm is a highly configurable open source workload manager. See the [Slurm proj
     8. [Capturing logs and configuration for troubleshooting](#capturing-logs-and-configuration-data-for-troubleshooting)
 6. [Contributing](#contributing)
 ---
-## Managing Slurm Clusters in 4.0.5
+## Managing Slurm Clusters in 4.0.6
 
 ### Making Cluster Changes
 In CycleCloud, cluster changes can be made using the "Edit" dialog from the cluster page in the GUI or from the CycleCloud CLI.   Cluster topology changes, such as new partitions, generally require editing and re-importing the cluster template.   This can be applied to live, running clusters as well as terminated clusters.   It is also possible to import changes as a new Template for future cluster creation via the GUI.

azure-slurm-install/setup.py

@@ -7,7 +7,7 @@
 from setuptools.command.test import Command
 from setuptools.command.test import test as TestCommand  # noqa: N812
 
-__version__ = "4.0.5"
+__version__ = "4.0.6"
 CWD = os.path.dirname(os.path.abspath(__file__))
 
 

azure-slurm-install/start-services.sh

@@ -217,6 +217,11 @@ ensure_enroot_dir() {
     #
     CONF=/etc/enroot/enroot.conf
 
+    # no-op if enroot.conf does not exist
+    if [ ! -f "$CONF" ]; then
+        return 0
+    fi
+
     # extract ENROOT_TEMP_PATH value
     ENROOT_TEMP_PATH=$(awk '$1=="ENROOT_TEMP_PATH"{print $2}' "$CONF")
 

azure-slurm-install/templates/enroot.conf.template

@@ -6,4 +6,6 @@ ENROOT_SQUASH_OPTIONS -noI -noD -noF -noX -no-duplicates
 ENROOT_MOUNT_HOME y
 ENROOT_RESTRICT_DEV y
 ENROOT_ROOTFS_WRITABLE y
-MELLANOX_VISIBLE_DEVICES all
+MELLANOX_VISIBLE_DEVICES all
+# Trailing comment - this file requires at least one trailing newline after the above
+

azure-slurm/setup.py

@@ -7,7 +7,7 @@
 from setuptools.command.test import Command
 from setuptools.command.test import test as TestCommand  # noqa: N812
 
-__version__ = "4.0.5"
+__version__ = "4.0.6"
 CWD = os.path.dirname(os.path.abspath(__file__))
 
 

azure-slurm/slurmcc/cli.py

@@ -39,7 +39,7 @@
 from . import topology
 
 
-VERSION = "4.0.5"
+VERSION = "4.0.6"
 
 
 def csv_list(x: str) -> List[str]:

project.ini

@@ -1,11 +1,11 @@
 [project]
 name = slurm
 label = Slurm
-version = 4.0.5
+version = 4.0.6
 type = scheduler
 
 [blobs]
-Files = azure-slurm-pkg-4.0.5.tar.gz, azure-slurm-install-pkg-4.0.5.tar.gz
+Files = azure-slurm-pkg-4.0.6.tar.gz, azure-slurm-install-pkg-4.0.6.tar.gz
 
 [spec scheduler]
 run_list = role[slurm_scheduler_role]

specs/default/chef/site-cookbooks/slurm/attributes/default.rb

@@ -1,6 +1,6 @@
 # Copyright (c) Microsoft Corporation. All rights reserved.
 # Licensed under the MIT License.
-default[:slurm][:autoscale_version] = "4.0.5"
+default[:slurm][:autoscale_version] = "4.0.6"
 default[:slurm][:version] = "23.11.9-1"
 default[:slurm][:user][:name] = 'slurm'
 default[:slurm][:cyclecloud_api] = "cyclecloud_api-8.4.1-py2.py3-none-any.whl"

specs/default/chef/site-cookbooks/slurm/metadata.rb

@@ -4,7 +4,7 @@
 license 'All Rights Reserved'
 description 'Installs/Configures slurm'
 long_description 'Installs/Configures slurm'
-version '4.0.5'
+version '4.0.6'
 chef_version '>= 12.1' if respond_to?(:chef_version)
 
 %w{ cuser cshared }.each {|c| depends c}

specs/scheduler/cluster-init/scripts/00-install.sh

@@ -13,13 +13,20 @@ find_python3() {
         return 0
     fi
     for version in $( seq 11 20 ); do
-        which python3.$version
+        which python3.$version > /dev/null 2>/dev/null
         if [ $? == 0 ]; then
-            return 0
+            python3.$version -m "import yaml, venv"
+            if [ $? == 0 ]; then
+                # write to stdout the validated path
+                which python3.$version
+                return 0
+            else
+                echo Warning: Found python3.$version but venv and/or yaml are not installed. 1>&2
+            fi
         fi
     done
-    echo Could not find python3 version 3.11 >&2
-    return 1
+    # Quietly return nothing
+    return 0
 }
 
 install_python3() {
@@ -29,6 +36,7 @@ install_python3() {
         export PYTHON_BIN
         return 0
     fi
+    echo "No suitable python3 installation found, beginning installation." >&2
     # NOTE: based off of healthagent 00-install.sh, but we have different needs - we don't need the devel/systemd paths.
     # most likely if healthagent is already installed, this won't be an issue.
     if [ -f /etc/os-release ]; then