kudos/check_health
kudos/check_health
2
1
Fork 0
Code Issues 7 Pull Requests Projects Releases 1 Wiki Activity

Added possbility to do checks based on weekday in check_exadata_zfs_share_replication plugin

Browse Source 
Added data_has_newline()
This commit is contained in:
Patrick Van der Veken 2020-01-30 11:43:36 +01:00
parent 9fcf788b59
commit 982a237d77
5 changed files with 116 additions and 66 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
etc/opt/hc/check_exadata_zfs_share_replication.conf.dist
View File

@ -33,21 +33,22 @@ max_replication_lag=300
# will for this given share will be skipped (this allows for exclusion of shares)
# In order to check share(s) for a given ZFS appliance at least one configuration
# entry must be present: either a wildcard or custom entry.
# Caveat: any share must finally resolve to one entry only.
# Format:
# zfs:<host_name>:<replication_name|*>:<true|false|*>:<success|failed|*>:[<max_replication_lag>]
# zfs:<host_name>:<replication_name|*>:<true|false|*>:<success|failed|*>:[<max_replication_lag>]:[day1,day2,..|*>]
# Examples:
# check rep_share1 on myzfs1 with a custom threshold of 300 seconds
# zfs:myzfs1:rep_share1:*:*:600
# check all shares of myzfs2 with a custom threshold of 1200 seconds
# zfs:myzfs2:*:*:*:1200
# check all shares of myzfs3 with the general threshold
# zfs:myzfs3:*:*:*:
# check rep_share1 on myzfs1 with a custom threshold of 300 seconds on every day of the week
# zfs:myzfs1:rep_share1:*:*:600:*
# check all shares of myzfs2 with a custom threshold of 1200 seconds on Sunday and Monday
# zfs:myzfs2:*:*:*:1200:Sun,Mon
# check all shares of myzfs3 with the general threshold but only on Friday
# zfs:myzfs3:*:*:*:Fri
# disable all shares of myzfs4 from checking
# zfs:myzfs4:*:*:*:0
# zfs:myzfs4:*:*:*:0:*
# disable check of rep_share7 on myzfs5
# zfs:myzfs5:rep_share7:*:*:0
# check that rep_share4 on myzfs6 is inactive
# zfs:myzfs6:rep_share4:false:*:
# zfs:myzfs5:rep_share7:*:*:0:*
# check that rep_share4 on myzfs6 is inactive (every day of the week)
# zfs:myzfs6:rep_share4:false:*:*
#******************************************************************************

1
etc/opt/hc/check_exadata_zfs_share_usage.conf.dist
View File

@ -33,6 +33,7 @@ max_space_usage=90
# will for this given share will be skipped (this allows for exclusion of shares)
# In order to check share(s) for a given ZFS appliance at least one configuration
# entry must be present: either a wildcard or custom entry.
# Caveat: any share must finally resolve to one entry only.
# Format:
# zfs:<host_name>:<project_name|*>:<share_name|*>:[<usage_threshold_%>]
# Examples:

26
opt/hc/lib/core/include_data.sh
View File

@ -30,7 +30,7 @@
# RETURNS: 0
function version_include_data
{
typeset _VERSION="2019-07-14" # YYYY-MM-DD
typeset _VERSION="2020-01-27" # YYYY-MM-DD
print "INFO: $0: ${_VERSION#version_*}"
@ -196,10 +196,30 @@ done
return 0
}
# -----------------------------------------------------------------------------
# @(#) FUNCTION: data_has_newline()
# DOES: checks if a string contains newlines
# EXPECTS: $1=haystack [string]
# OUTPUTS: n/a
# RETURNS: 0=no newline found; 1=newlines found
# REQUIRES: n/a
function data_has_newline
{
(( ARG_DEBUG > 0 && ARG_DEBUG_LEVEL > 0 )) && set "${DEBUG_OPTS}"
typeset _HAYSTACK="${1}"
typeset _COUNT=$(print -R "${_HAYSTACK}" | wc -l 2>/dev/null)
(( _COUNT > 1 )) && return 1
return 0
}
# -----------------------------------------------------------------------------
# @(#) FUNCTION: data_magic_quote()
# DOES: magically quotes a needle in a string (default needle is: %)
# EXPECTS: $1=to be magically quoted [string]; $2=needle [string]
# EXPECTS: to be magically quoted [string]; $2=needle [string]
# OUTPUTS: magically quoted [string]
# RETURNS: n/a
# REQUIRES: n/a
@ -217,7 +237,7 @@ return 0
# -----------------------------------------------------------------------------
# @(#) FUNCTION: data_magic_unquote()
# DOES: magically unquotes a needle in a string (default needle is: %)
# EXPECTS: $1=to be magically unquoted [string]; $2=needle [string]
# EXPECTS: to be magically unquoted [string]; $2=needle [string]
# OUTPUTS: magically unquoted [string]
# RETURNS: n/a
# REQUIRES: n/a

116
opt/hc/lib/platform/exadata/check_exadata_zfs_share_replication.sh
View File

@ -20,7 +20,8 @@
# DOES: see _show_usage()
# EXPECTS: see _show_usage()
# REQUIRES: data_comma2space(), data_contains_string(), data_get_lvalue_from_config(),
# dump_logs(), init_hc(), linux_exec_ssh(), log_hc(), warn()
# data_has_newline(), dump_logs(), init_hc(), linux_exec_ssh(),
# log_hc(), warn()
#
# @(#) HISTORY:
# @(#) 2019-02-18: initial version [Patrick Van der Veken]
@ -28,7 +29,8 @@
# @(#) 2019-03-16: replace 'which' [Patrick Van der Veken]
# @(#) 2019-04-12: small fixes [Patrick Van der Veken]
# @(#) 2019-05-14: small fixes [Patrick Van der Veken]
# @(#) 2019-07-05: help fix [Patrick Van der Veken]
# @(#) 2020-01-27: addition of day check option +
# @(#) newline config value check [Patrick Van der Veken]
# -----------------------------------------------------------------------------
# DO NOT CHANGE THIS FILE UNLESS YOU KNOW WHAT YOU ARE DOING!
#******************************************************************************
@ -38,7 +40,7 @@ function check_exadata_zfs_share_replication
{
# ------------------------- CONFIGURATION starts here -------------------------
typeset _CONFIG_FILE="${CONFIG_DIR}/$0.conf"
typeset _VERSION="2019-07-05" # YYYY-MM-DD
typeset _VERSION="2020-01-27" # YYYY-MM-DD
typeset _SUPPORTED_PLATFORMS="Linux" # uname -s match
# replication query script -- DO NOT CHANGE --
# prj1/share1:true:idle:success:111
@ -75,12 +77,14 @@ typeset _CFG_SSH_USER=""
typeset _CFG_ZFS_HOSTS=""
typeset _CFG_ZFS_HOST=""
typeset _CFG_ZFS_LINE=""
typeset _CFG_REPLICATION_DAYS=""
typeset _REPLICATION_ENABLED=""
typeset _REPLICATION_LAG=""
typeset _REPLICATION_RESULT=""
typeset _SSH_BIN=""
typeset _SSH_OUTPUT=""
typeset _ZFS_DATA=""
typeset _WEEKDAY=$(data_lc "$(date '+%a' 2>/dev/null)") # Sun
# handle arguments (originally comma-separated)
for _ARG in ${_ARGS}
@ -210,6 +214,7 @@ do
_CFG_REPLICATION_ENABLED=""
_CFG_REPLICATION_RESULT=""
_CFG_REPLICATION_LAG=""
_CFG_REPLICATION_DAYS=""
# which values to use (general or custom?), keep in mind wildcards (custom will overrule wildcard entry)
_CFG_ZFS_LINE=$(grep -E -e "^zfs:${_ZFS_HOST}:[*]:" ${_CONFIG_FILE} 2>/dev/null)
@ -219,6 +224,7 @@ do
_CFG_REPLICATION_ENABLED=$(print "${_CFG_ZFS_LINE}" | cut -f4 -d':' 2>/dev/null)
_CFG_REPLICATION_RESULT=$(print "${_CFG_ZFS_LINE}" | cut -f5 -d':' 2>/dev/null)
_CFG_REPLICATION_LAG=$(print "${_CFG_ZFS_LINE}" | cut -f6 -d':' 2>/dev/null)
_CFG_REPLICATION_DAYS=$(print "${_CFG_ZFS_LINE}" | cut -f7 -d':' 2>/dev/null)
# null value means general threshold
if [[ -z "${_CFG_REPLICATION_LAG}" ]]
then
@ -229,10 +235,18 @@ do
_CFG_ZFS_LINE=$(grep -E -e "^zfs:${_ZFS_HOST}:${_REPLICATION_NAME}:" ${_CONFIG_FILE} 2>/dev/null)
if [[ -n "${_CFG_ZFS_LINE}" ]]
then
data_has_newline "${_CFG_ZFS_LINE}"
# shellcheck disable=SC2181
if (( $? > 0 ))
then
warn "ignoring ${_ZFS_HOST}:${_REPLICATION_NAME} because it parses to multiple results in ${_CONFIG_FILE}"
continue
fi
(( ARG_DEBUG > 0 )) && debug "found custom definition for ${_ZFS_HOST}:${_REPLICATION_NAME} in configuration file ${_CONFIG_FILE}"
_CFG_REPLICATION_ENABLED=$(print "${_CFG_ZFS_LINE}" | cut -f4 -d':' 2>/dev/null)
_CFG_REPLICATION_RESULT=$(print "${_CFG_ZFS_LINE}" | cut -f5 -d':' 2>/dev/null)
_CFG_REPLICATION_LAG=$(print "${_CFG_ZFS_LINE}" | cut -f6 -d':' 2>/dev/null)
_CFG_REPLICATION_DAYS=$(print "${_CFG_ZFS_LINE}" | cut -f7 -d':' 2>/dev/null)
# null value means general threshold
if [[ -z "${_CFG_REPLICATION_LAG}" ]]
then
@ -262,56 +276,65 @@ do
# fixed defaults if missing
[[ -z "${_CFG_REPLICATION_ENABLED}" || "${_CFG_REPLICATION_ENABLED}" = '*' ]] && _CFG_REPLICATION_ENABLED="true"
[[ -z "${_CFG_REPLICATION_RESULT}" || "${_CFG_REPLICATION_RESULT}" = '*' ]] && _CFG_REPLICATION_RESULT="success"
_CFG_REPLICATION_DAYS=$(data_lc "${_CFG_REPLICATION_DAYS}")
[[ -z "${_CFG_REPLICATION_DAYS}" || "${_CFG_REPLICATION_DAYS}" = '*' ]] && _CFG_REPLICATION_DAYS="${_WEEKDAY}"
# perform checks
# check replication enabled state (active or not?)
if [[ $(data_lc "${_REPLICATION_ENABLED}") != $(data_lc "${_CFG_REPLICATION_ENABLED}") ]]
then
_MSG="state for ${_ZFS_HOST}:${_REPLICATION_NAME} is NOK [${_REPLICATION_ENABLED}!=${_CFG_REPLICATION_ENABLED}]"
_STC=1
else
_MSG="state for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_ENABLED}==${_CFG_REPLICATION_ENABLED}]"
_STC=0
fi
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_ENABLED}" "${_CFG_REPLICATION_ENABLED}"
fi
# check replication last result (success or not?)
if [[ $(data_lc "${_REPLICATION_RESULT}") != $(data_lc "${_CFG_REPLICATION_RESULT}") ]]
then
_MSG="result for ${_ZFS_HOST}:${_REPLICATION_NAME} is NOK [${_REPLICATION_RESULT}!=${_CFG_REPLICATION_RESULT}]"
_STC=1
else
_MSG="result for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_RESULT}==${_CFG_REPLICATION_RESULT}]"
_STC=0
fi
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_RESULT}" "${_CFG_REPLICATION_RESULT}"
fi
# check replication lag
# caveat: replication lag is <unknown> at initial replication
data_contains_string "${_REPLICATION_LAG}" "unknown"
# shellcheck disable=SC2181
# do we need to perform the check today?
data_contains_string "${_CFG_REPLICATION_DAYS}" "${_WEEKDAY}"
if (( $? > 0 ))
then
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is unknown"
_REPLICATION_LAG=-1
_STC=1
else
if (( _REPLICATION_LAG > _CFG_REPLICATION_LAG ))
# check replication enabled state (active or not?)
if [[ $(data_lc "${_REPLICATION_ENABLED}") != $(data_lc "${_CFG_REPLICATION_ENABLED}") ]]
then
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is too big [${_REPLICATION_LAG}>${_CFG_REPLICATION_LAG}]"
_MSG="state for ${_ZFS_HOST}:${_REPLICATION_NAME} is NOK [${_REPLICATION_ENABLED}!=${_CFG_REPLICATION_ENABLED}]"
_STC=1
else
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_LAG}<=${_CFG_REPLICATION_LAG}]"
_MSG="state for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_ENABLED}==${_CFG_REPLICATION_ENABLED}]"
_STC=0
fi
fi
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_LAG}" "${_CFG_REPLICATION_LAG}"
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_ENABLED}" "${_CFG_REPLICATION_ENABLED}"
fi
# check replication last result (success or not?)
if [[ $(data_lc "${_REPLICATION_RESULT}") != $(data_lc "${_CFG_REPLICATION_RESULT}") ]]
then
_MSG="result for ${_ZFS_HOST}:${_REPLICATION_NAME} is NOK [${_REPLICATION_RESULT}!=${_CFG_REPLICATION_RESULT}]"
_STC=1
else
_MSG="result for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_RESULT}==${_CFG_REPLICATION_RESULT}]"
_STC=0
fi
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_RESULT}" "${_CFG_REPLICATION_RESULT}"
fi
# check replication lag
# caveat: replication lag is <unknown> at initial replication
data_contains_string "${_REPLICATION_LAG}" "unknown"
# shellcheck disable=SC2181
if (( $? > 0 ))
then
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is unknown"
_REPLICATION_LAG=-1
_STC=1
else
if (( _REPLICATION_LAG > _CFG_REPLICATION_LAG ))
then
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is too big [${_REPLICATION_LAG}>${_CFG_REPLICATION_LAG}]"
_STC=1
else
_MSG="lag for ${_ZFS_HOST}:${_REPLICATION_NAME} is OK [${_REPLICATION_LAG}<=${_CFG_REPLICATION_LAG}]"
_STC=0
fi
fi
if (( _LOG_HEALTHY > 0 || _STC > 0 ))
then
log_hc "$0" ${_STC} "${_MSG}" "${_REPLICATION_LAG}" "${_CFG_REPLICATION_LAG}"
fi
else
warn "check of ${_ZFS_HOST}:${_REPLICATION_NAME} was not configured for today"
fi
done
@ -328,11 +351,10 @@ CONFIG : $3 with parameters:
log_healthy=<yes|no>
ssh_user=<ssh_user_account>
ssh_key_file=<ssh_private_key_file>
ssh_opts=<ssh_options>
max_replication_lag=<general_max_replication>
and formatted stanzas of:
zfs:<host_name>:<replication_name>:<replication_enabled>:<replication_result>:<max_replication_lag>
PURPOSE : Checks the replication state, sync status and maximum lag of the configured ZFS hosts/shares
zfs:<host_name>:<replication_name>:<replication_enabled>:<replication_result>:<max_replication_lag>:<day1,day2>
PURPOSE : Checks the replication state, sync status and maximum lag of the configured ZFS hosts/shares on certain days
CLI: zfs > shares > replications > packages > select (action) > show
LOG HEALTHY : Supported

16
opt/hc/lib/platform/exadata/check_exadata_zfs_share_usage.sh
View File

@ -19,8 +19,8 @@
# @(#) MAIN: check_exadata_zfs_share_usage
# DOES: see _show_usage()
# EXPECTS: see _show_usage()
# REQUIRES: data_comma2space(), data_get_lvalue_from_config, dump_logs(),
# init_hc(), linux_exec_ssh(), log_hc(), warn()
# REQUIRES: data_comma2space(), data_get_lvalue_from_config, data_has_newline(),
# dump_logs(), init_hc(), linux_exec_ssh(), log_hc(), warn()
#
# @(#) HISTORY:
# @(#) 2019-02-18: initial version [Patrick Van der Veken]
@ -28,7 +28,7 @@
# @(#) 2019-04-09: fix bad math in ZFS script & HC message [Patrick Van der Veken]
# @(#) 2019-04-12: small fixes [Patrick Van der Veken]
# @(#) 2019-05-14: small fixes [Patrick Van der Veken]
# @(#) 2019-07-05: help fix [Patrick Van der Veken]
# @(#) 2020-01-27: newline config value check [Patrick Van der Veken]
# -----------------------------------------------------------------------------
# DO NOT CHANGE THIS FILE UNLESS YOU KNOW WHAT YOU ARE DOING!
#******************************************************************************
@ -38,7 +38,7 @@ function check_exadata_zfs_share_usage
{
# ------------------------- CONFIGURATION starts here -------------------------
typeset _CONFIG_FILE="${CONFIG_DIR}/$0.conf"
typeset _VERSION="2019-07-05" # YYYY-MM-DD
typeset _VERSION="2020-01-27" # YYYY-MM-DD
typeset _SUPPORTED_PLATFORMS="Linux" # uname -s match
# usage query script -- DO NOT CHANGE --
# prj1:share1:16
@ -233,6 +233,13 @@ do
_CFG_ZFS_LINE=$(grep -E -e "^zfs:${_ZFS_HOST}:${_PROJECT_NAME}:${_SHARE_NAME}:" ${_CONFIG_FILE} 2>/dev/null)
if [[ -n "${_CFG_ZFS_LINE}" ]]
then
data_has_newline "${_CFG_ZFS_LINE}"
# shellcheck disable=SC2181
if (( $? > 0 ))
then
warn "ignoring ${_ZFS_HOST}:${_PROJECT_NAME}:${_SHARE_NAME} because it parses to multiple results in ${_CONFIG_FILE}"
continue
fi
(( ARG_DEBUG > 0 )) && debug "found custom definition for ${_ZFS_HOST}:${_PROJECT_NAME}/${_SHARE_NAME} in configuration file ${_CONFIG_FILE}"
_CFG_SPACE_THRESHOLD=$(print "${_CFG_ZFS_LINE}" | cut -f5 -d':' 2>/dev/null)
# null value means general threshold
@ -290,7 +297,6 @@ CONFIG : $3 with parameters:
log_healthy=<yes|no>
ssh_user=<ssh_user_account>
ssh_key_file=<ssh_private_key_file>
ssh_opts=<ssh_options>
max_space_usage=<general_max_space_treshold>
and formatted stanzas of:
zfs:<host_name>:<project_name>:<share_name>:<max_space_threshold>