diff options
Diffstat (limited to 'puppet/modules/site_check_mk/files')
-rwxr-xr-x | puppet/modules/site_check_mk/files/agent/local_checks/couchdb/leap_couch_stats.sh | 106 | ||||
-rw-r--r-- | puppet/modules/site_check_mk/files/extra_service_conf.mk | 8 |
2 files changed, 108 insertions, 6 deletions
diff --git a/puppet/modules/site_check_mk/files/agent/local_checks/couchdb/leap_couch_stats.sh b/puppet/modules/site_check_mk/files/agent/local_checks/couchdb/leap_couch_stats.sh new file mode 100755 index 00000000..51a8ac52 --- /dev/null +++ b/puppet/modules/site_check_mk/files/agent/local_checks/couchdb/leap_couch_stats.sh @@ -0,0 +1,106 @@ +#!/bin/bash +# +# todo: +# - thresholds +# - couch response time +# - make CURL/URL/DBLIST_EXCLUDE vars configurable +# - move load_nagios_utils() to helper library so we can use it from multiple scripts + +start_time=$(date +%s.%N) + +CURL='curl -s --netrc-file /etc/couchdb/couchdb.netrc' +URL='http://127.0.0.1:5984' +TMPFILE=$(mktemp) +DBLIST_EXCLUDE='user-' +PREFIX='Couchdb_' + + +load_nagios_utils () { + # load the nagios utils + # in debian, the package nagios-plugins-common installs utils.sh to /usr/lib/nagios/plugins/utils.sh + utilsfn= + for d in $PROGPATH /usr/lib/nagios/plugins /usr/lib64/nagios/plugins /usr/local/nagios/libexec /opt/nagios-plugins/libexec . ; do + if [ -f "$d/utils.sh" ]; then + utilsfn=$d/utils.sh; + fi + done + if [ "$utilsfn" = "" ]; then + echo "UNKNOWN - cannot find utils.sh (part of nagios plugins)"; + exit 3; + fi + . "$utilsfn"; + STATE[$STATE_OK]='OK' + STATE[$STATE_WARNING]='Warning' + STATE[$STATE_CRITICAL]='Critical' + STATE[$STATE_UNKNOWN]='Unknown' + STATE[$STATE_DEPENDENT]='Dependend' +} + +get_global_stats_perf () { + trap "localexit=3" ERR + local localexit db_count + localexit=0 + + # get a list of all dbs + $CURL -X GET $URL/_all_dbs | json_pp | egrep -v '(\[|\])' > $TMPFILE + + db_count=$( wc -l < $TMPFILE) + excluded_db_count=$( grep -c "$DBLIST_EXCLUDE" $TMPFILE ) + + echo "db_count=$db_count|excluded_db_count=$excluded_db_count" + return ${localexit} +} + +db_stats () { + trap "localexit=3" ERR + local db db_stats doc_count del_doc_count localexit + localexit=0 + + db=$1 + perf="$perf|${db}_docs=$( $CURL -s -X GET ${URL}/$db | json_pp |grep 'doc_count' | sed 's/[^0-9]//g' )" + db_stats=$( $CURL -s -X GET ${URL}/$db | json_pp ) + + doc_count=$( echo "$db_stats" | grep 'doc_count' | grep -v 'deleted_doc_count' | sed 's/[^0-9]//g' ) + del_doc_count=$( echo "$db_stats" | grep 'doc_del_count' | sed 's/[^0-9]//g' ) + + # don't divide by zero + if [ $del_doc_count -eq 0 ] + then + del_doc_perc=0 + else + del_doc_perc=$(( del_doc_count * 100 / doc_count )) + fi + + bytes=$( echo "$db_stats" | grep disk_size | sed 's/[^0-9]//g' ) + disk_size=$( echo "scale = 2; $bytes / 1024 / 1024" | bc -l ) + + echo -n "${localexit} ${PREFIX}${db}_database ${db}_docs=$doc_count|${db}_deleted_docs=$del_doc_count|${db}_deleted_docs_percentage=${del_doc_perc}%" + printf "|${db}_disksize_mb=%02.2fmb ${STATE[localexit]}: database $db\n" "$disk_size" + + return ${localexit} +} + +# main + +load_nagios_utils + +# per-db stats +# get a list of all dbs +$CURL -X GET $URL/_all_dbs | json_pp | egrep -v '(\[|\])' > $TMPFILE + +# get list of dbs to check +dbs=$( grep -v "${DBLIST_EXCLUDE}" $TMPFILE | tr -d '\n"' | sed 's/,/ /g' ) + +for db in $dbs +do + db_stats "$db" +done + +# show global couchdb stats +global_stats_perf=$(get_global_stats_perf) +exitcode=$? + +end_time=$(date +%s.%N) +duration=$( echo "scale = 2; $end_time - $start_time" | bc -l ) + +printf "${exitcode} ${PREFIX}global_stats ${global_stats_perf}|script_duration=%02.2fs ${STATE[exitcode]}: global couchdb status\n" "$duration" diff --git a/puppet/modules/site_check_mk/files/extra_service_conf.mk b/puppet/modules/site_check_mk/files/extra_service_conf.mk index 61b0dd39..a4c6e769 100644 --- a/puppet/modules/site_check_mk/files/extra_service_conf.mk +++ b/puppet/modules/site_check_mk/files/extra_service_conf.mk @@ -5,14 +5,10 @@ extra_service_conf["max_check_attempts"] = [ ] # -# run check_mk_agent every 2 minutes if it terminates successfully. +# run check_mk_agent every 10 minutes if it terminates successfully. # see https://leap.se/code/issues/6539 for the rationale # -# update: temporarily set interval to 60 minutes until we solve the -# issue with the users db getting bloated with deleted -# test users. -# extra_service_conf["normal_check_interval"] = [ - ("60", ALL_HOSTS , "Check_MK" ) + ("10", ALL_HOSTS , "Check_MK" ) ] |