Nodeletd Restart Failing At Phase 2

Problem

• Restarting nodeletd fails on phase 2 with below error:

None

--- /opt/pf9/pf9-kube/worker_scripts/040-docker_configure.sh start at 2021-05-25 07:43:55 ---
[2021-05-25 07:43:55] docker-ce-19.03.11-3.el7.x86_64
[2021-05-25 07:43:55] docker-ce-cli-19.03.11-3.el7.x86_64
[2021-05-25 07:43:55] configure_docker_storage: deriving configuration from /etc/sysconfig/docker-storage
[2021-05-25 07:43:55] os_centos.sh: line 78: /opt/pf9/python/pf9/python: No such file or directory
Error: exit status 127
Exit Status: 127

Environment

• Platform9 Managed Kubernetes - v5.2 and below version

• Nodeletd

Cause

The /opt/pf9/pf9-kube/os_centos.sh script contains a typo for python path:

os_centos.sh

local storage_cfg_backup=/etc/sysconfig/docker-storage.rpmsave
    if [ ! -f $storage_cfg ] && [ -f $storage_cfg_backup ] ; then
        echo "configure_docker_storage: restoring $storage_cfg"
        mv -f $storage_cfg_backup $storage_cfg
    fi    # Ensure backward compatibility with hosts where docker storage
    # is already configured
    if [ -f $storage_cfg ]; then
        echo "configure_docker_storage: deriving configuration from $storage_cfg"
        local PYTHON=*/opt/pf9/python/pf9/python*

Resolution

• Correct python path from /opt/pf9/python/pf9/python to /opt/pf9/python/bin/python of /opt/pf9/pf9-kube/os_centos.sh script

• The issue is fixed from Platform9 Managed Kubernetes v5.3 onwards.