f87df98a96
Several collect plugins were using the 'sm-query service' with
'management-ip' as the target service to determine if the host
collect is running on is the active controller.
sm-query service management-ip
Unfortunately, a recent update changed the name of the 'management-ip'
service. This change lead to that query always returning a 'disabled'
status.
This update changes to use 'sm-query service-group controller-services'
instead of a specific service query that might change again in the
future. It also migrates the replicated is_service_active function to
the collect_utils file making it common and reusable by all aspects of
collect. This function was renamed to 'is_controller_active' because
it's no longer representing the activity state of a particular service.
Test Plan:
PASS: Verify 'sysinv' plugin creates var/extra/inventory.info file
on active controller. sysinv CLI command output.
PASS: Verify 'fm' plugin creates var/extra/alarm.info file
on active controller. FM alarm and event listings.
PASS: Verify 'ceph' plugin creates var/extra/ceph.info file
on active controller with 'ceph df' output content.
PASS: Verify 'dc' plugin creates var/extra/distributed_cloud.info file
on active controller of a system controller or subcloud.
Closes-Bug: 2070496
Change-Id: I9989708f1d87a5ef312129cfe3ede8c862764cb0
Signed-off-by: Eric MacDonald <eric.macdonald@windriver.com>
73 lines
2.0 KiB
Bash
Executable File
73 lines
2.0 KiB
Bash
Executable File
#! /bin/bash
|
|
#
|
|
# Copyright (c) 2013-2014 Wind River Systems, Inc.
|
|
#
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
#
|
|
|
|
|
|
# Loads Up Utilities and Commands Variables
|
|
source /usr/local/sbin/collect_parms
|
|
source /usr/local/sbin/collect_utils
|
|
|
|
SERVICE="ceph"
|
|
LOGFILE="${extradir}/ceph.info"
|
|
echo "${hostname}: Ceph Info .........: ${LOGFILE}"
|
|
|
|
function exit_if_timeout {
|
|
if [ "$?" = "124" ] ; then
|
|
echo "Exiting due to ceph command timeout" >> ${LOGFILE}
|
|
exit 0
|
|
fi
|
|
}
|
|
|
|
###############################################################################
|
|
# Only Controller
|
|
###############################################################################
|
|
if [ "$nodetype" = "controller" ] ; then
|
|
|
|
# Using timeout with all ceph commands because commands can hang for
|
|
# minutes if the ceph cluster is down. If ceph is not configured, the
|
|
# commands return immediately.
|
|
|
|
delimiter ${LOGFILE} "ceph status"
|
|
timeout 30 ceph status >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph mon dump"
|
|
timeout 30 ceph mon dump >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph osd dump"
|
|
timeout 30 ceph osd dump >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph osd tree"
|
|
timeout 30 ceph osd tree >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph osd crush dump"
|
|
timeout 30 ceph osd crush dump >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
is_active_controller
|
|
if [ "$?" = "0" ] ; then
|
|
exit 0
|
|
fi
|
|
|
|
delimiter ${LOGFILE} "ceph df"
|
|
timeout 30 ceph df >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph osd df tree"
|
|
timeout 30 ceph osd df tree >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
delimiter ${LOGFILE} "ceph health detail"
|
|
timeout 30 ceph health detail >> ${LOGFILE} 2>>${COLLECT_ERROR_LOG}
|
|
exit_if_timeout
|
|
|
|
fi
|
|
|
|
exit 0
|