generic checks

This commit is contained in:
Martin Berghaus
2025-04-02 17:28:34 +02:00
parent c15723b499
commit 5cc79adc0b
6 changed files with 131 additions and 45 deletions

View File

@@ -0,0 +1,68 @@
#!/bin/bash
_REMOTE_HOST="${1:?"FQDN of server missing: e.g. host.example.net[:port]"}"
_REMOTE_HOSTNAME_FQDN="${_REMOTE_HOST%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_HOSTNAME_SHORT="${_REMOTE_HOSTNAME_FQDN%%.*}" #Removes longest matching pattern '.*' from the end
_REMOTE_PORT="${_REMOTE_HOST}:"
_REMOTE_PORT="${_REMOTE_PORT#*:}" #Removes shortest matching pattern '*:' from the begin
_REMOTE_PORT="${_REMOTE_PORT%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_PORT="${_REMOTE_PORT:-"22"}"
_REMOTE_USER="monitoring"
_SOCKET='~/.ssh/%r@%h:%p'
function checkOrStartSSHMaster() {
timeout --preserve-status 1 "ssh -O check -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN}" &> /dev/null \
&& echo "master checked" \
&& return 0
ssh -O stop -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} &> /dev/null
ssh -o ControlMaster=auto \
-o ControlPath=${_SOCKET} \
-o ControlPersist=65 \
-p ${_REMOTE_PORT} \
-f ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} exit &> /dev/null \
&& return 0
echo "Fail: checkOrStartMaster()"
return 1
}
function checkViaHTTP() {
_STATUS="$(curl -I http://${_REMOTE_HOSTNAME_FQDN} 2>/dev/null | head -n 1 | cut -d$' ' -f2)"
[ "${_STATUS}" == "200" ] \
&& echo "OK" \
&& return 0
return 1
}
function checkViaHTTPS() {
_STATUS="$(curl -k -I https://${_REMOTE_HOSTNAME_FQDN} 2>/dev/null | head -n 1 | cut -d$' ' -f2)"
[ "${_STATUS}" == "200" ] \
&& echo "OK" \
&& return 0
return 1
}
#grep:
# -E Use regexp, '.*' => any chars between 'Active:' and '(running)', the round brackets are escaped.
#cut:
# -d Delimiter, marker where to cut (here ;)
# -f Index of column to show (One based, so there is no -f0)
function checkViaSSH() {
checkOrStartSSHMaster \
|| return 1
_RESULT=$(ssh -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} 'systemctl status nginx.service' | grep -E 'Active:.*\(running\)' | cut -d';' -f2)
! [ -z "${_RESULT}" ] && echo "OK#UPTIME:${_RESULT}" || echo "FAIL"
}
#checkViaHTTP && exit 0
#checkViaHTTPS && exit 0
checkViaSSH && exit 0
exit 1

View File

@@ -0,0 +1,9 @@
#!/bin/bash
_SERVER="${1:?"FQDN of server missing"}"
# -4 Use IPv4
# -W SECONDS Wait seconds for an answer
# -c COUNT_VALUE Count of pings being executed
_RESULT="$(ping -4 -W 1 -c 1 "${_SERVER}" | grep "time=" | cut -d'=' -f4)"
! [ -z "${_RESULT}" ] && echo "OK#RTT: ${_RESULT}" || echo "FAIL#PLEASE USE FALLBACK!"

View File

@@ -0,0 +1,17 @@
#!/bin/bash
_URL="${1:?"URL of site missing"}"
#curl:
# --connect-timeout SECONDS Maximum time allowed for connection
# -k Allow connections to SSL sites without certs (H)
# -L Follow redirects (H)
# --max-time SECONDS Maximum time allowed for the transfer
# -s Silent mode. Don't output anything
# --head Show head information only
# --no-progress-meter Clean output for grep
#grep:
# -q Quite, no output just status codes
# -F Interpret search term as plain text
((curl --connect-timeout 10 --max-time 10 -k -s --head --no-progress-meter "${_URL}" | grep -qF '200 OK') && echo OK) || echo FAIL

View File

@@ -0,0 +1,96 @@
#!/bin/bash
_SCRIPT="$(readlink -f "${0}" 2> /dev/null)"
# Folders always ends with an tailing '/'
_CIS_ROOT="${_SCRIPT%%/script/monitor/*}/" #Removes longest matching pattern '/script/monitor/*' from the end
_DOMAIN="$("${_CIS_ROOT:?"Missing CIS_ROOT"}core/printOwnDomain.sh")"
_COMPOSITIONS="${_CIS_ROOT:?"Missing CIS_ROOT"}definitions/${_DOMAIN:?"Missing DOMAIN"}/compositions/"
_REMOTE_HOST="${1:?"FQDN of server missing: e.g. host.example.net[:port]"}"
_MODE="${2:-"normal"}"
_REMOTE_HOSTNAME_FQDN="${_REMOTE_HOST%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_HOSTNAME_SHORT="${_REMOTE_HOSTNAME_FQDN%%.*}" #Removes longest matching pattern '.*' from the end
_REMOTE_PORT="${_REMOTE_HOST}:"
_REMOTE_PORT="${_REMOTE_PORT#*:}" #Removes shortest matching pattern '*:' from the begin
_REMOTE_PORT="${_REMOTE_PORT%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_PORT="${_REMOTE_PORT:-"22"}"
_REMOTE_USER="monitoring"
_SOCKET='~/.ssh/%r@%h:%p'
_NOW_UTC_UNIXTIME=$(date -u +%s)
_DEBUG_PATH="/tmp/monitor/"
function checkOrStartSSHMaster() {
timeout --preserve-status 1 "ssh -O check -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN}" &> /dev/null \
&& echo "master checked" \
&& return 0
ssh -O stop -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} &> /dev/null
ssh -o ControlMaster=auto \
-o ControlPath=${_SOCKET} \
-o ControlPersist=65 \
-p ${_REMOTE_PORT} \
-f ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} exit &> /dev/null \
&& return 0
echo "Fail: checkOrStartMaster()"
return 1
}
function checkSync() {
checkOrStartSSHMaster \
|| return 1
[ "${_MODE}" == "debug" ] \
&& mkdir -p "${_DEBUG_PATH}" > /dev/null \
&& echo "Now: ${_NOW_UTC_UNIXTIME}" > ${_DEBUG_PATH}SECONDS_BEHIND_${_REMOTE_HOSTNAME_FQDN}.txt
! [ -d "${_COMPOSITIONS:?"Missing COMPOSITIONS"}" ] \
&& echo "OK#no compositions" \
&& return 0
_SNAPSHOTS="$(ssh -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} zfs list -po creation,name -r -t snapshot zpool1/persistent | grep -F @SYNC_${_REMOTE_HOSTNAME_FQDN})"
[ "${_MODE}" == "debug" ] \
&& echo "${_SNAPSHOTS}" > ${_DEBUG_PATH}SNAPSHOTS_${_REMOTE_HOSTNAME_FQDN}.txt
[ -z "${_SNAPSHOTS}" ] \
&& echo "FAIL#no snapshots" \
&& return 1
echo "OK#Checks running"
for _COMPOSITION_PATH in ${_COMPOSITIONS}*; do
# Skip if remote host is not responsible for this container-composition
grep -vF "${_REMOTE_HOST}" "${_COMPOSITION_PATH}/zfssync-hosts" &> /dev/null \
&& continue;
_COMPOSITION_NAME="${_COMPOSITION_PATH##*/}" #Removes longest matching pattern '*/' from the begin
_LAST_SNAPSHOT_UNIXTIME="$(echo "${_SNAPSHOTS}" | grep ${_COMPOSITION_NAME} | tail -n 1 | cut -d' ' -f1)"
_SECONDS_BEHIND=$[ ${_NOW_UTC_UNIXTIME} - ${_LAST_SNAPSHOT_UNIXTIME} ]
[ "${_MODE}" == "debug" ] \
&& echo "${_LAST_SNAPSHOT_UNIXTIME} ${_COMPOSITION_NAME} on ${_REMOTE_HOSTNAME_FQDN} behind: ${_SECONDS_BEHIND}s" >> ${_DEBUG_PATH}SECONDS_BEHIND_${_REMOTE_HOSTNAME_FQDN}.txt
[ "${_SECONDS_BEHIND}" -lt 40 ] \
&& continue
[ "${_SECONDS_BEHIND}" -lt 60 ] \
&& echo "ZFSSYNC_of_${_REMOTE_HOSTNAME_SHORT}_LAGGING?WARN#${_COMPOSITION_NAME} ${_SECONDS_BEHIND}s" \
&& continue
echo "ZFSSYNC_of_${_REMOTE_HOSTNAME_SHORT}_LAGGING?FAIL#${_COMPOSITION_NAME} ${_SECONDS_BEHIND}s"
done
}
RESULTS="$(checkSync)"
[ "${_MODE}" == "debug" ] \
&& echo "$RESULTS" > ${_DEBUG_PATH}RESULTS_${_REMOTE_HOSTNAME_FQDN}.txt
echo "$RESULTS"

View File

@@ -0,0 +1,58 @@
#!/bin/bash
_REMOTE_HOST="${1:?"FQDN of server missing: e.g. host.example.net[:port]"}"
_REMOTE_HOSTNAME_FQDN="${_REMOTE_HOST%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_HOSTNAME_SHORT="${_REMOTE_HOSTNAME_FQDN%%.*}" #Removes longest matching pattern '.*' from the end
_REMOTE_PORT="${_REMOTE_HOST}:"
_REMOTE_PORT="${_REMOTE_PORT#*:}" #Removes shortest matching pattern '*:' from the begin
_REMOTE_PORT="${_REMOTE_PORT%%:*}" #Removes longest matching pattern ':*' from the end
_REMOTE_PORT="${_REMOTE_PORT:-"22"}"
_REMOTE_USER="monitoring"
_SOCKET='~/.ssh/%r@%h:%p'
function checkOrStartSSHMaster() {
timeout --preserve-status 1 "ssh -O check -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN}" &> /dev/null \
&& echo "master checked" \
&& return 0
ssh -O stop -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} &> /dev/null
ssh -o ControlMaster=auto \
-o ControlPath=${_SOCKET} \
-o ControlPersist=65 \
-p ${_REMOTE_PORT} \
-f ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} exit &> /dev/null \
&& return 0
echo "Fail: checkOrStartMaster()"
return 1
}
function testSpace(){
checkOrStartSSHMaster \
|| return 1
local _RESULT="$(ssh -S ${_SOCKET} -p ${_REMOTE_PORT} ${_REMOTE_USER}@${_REMOTE_HOSTNAME_FQDN} 'zpool list -H -o capacity,name')"
local _SPACE_USED=$(echo "${_RESULT}" | /usr/bin/tail -n 1 | /usr/bin/cut -f1)
local _POOL=$(echo "${_RESULT}" | /usr/bin/tail -n 1 | /usr/bin/cut -f2)
[ -z "${_SPACE_USED}" ] \
&& echo "FAIL#NO value" \
&& return 0
[ "${1:?"Missing OK_THRESHOLD"}" -ge "${_SPACE_USED%\%*}" ] \
&& echo "OK#${_SPACE_USED} used ${_POOL}." \
&& return 0
[ "${2:?"Missing INFO_THRESHOLD"}" -ge "${_SPACE_USED%\%*}" ] \
&& echo "INFO#${_SPACE_USED} already used ${_POOL}." \
&& return 0
echo "FAIL#${_SPACE_USED} used ${_POOL}!"
return 0
}
testSpace 80 90 && exit 0
exit 1