Merge branch '6010-check-pqsql' into 'master'

psql: show unknown if database connection fails

See merge request !90

check_psqlserver

@@ -14,6 +14,7 @@
 # https://www.postgresql.org/docs/current/monitoring-stats.html
 # ----------------------------------------------------------------------
 # 2023-06-07  v0.2  <axel.hahn@unibe.ch>
+# 2023-06-08  v0.3  <axel.hahn@unibe.ch>  show unknown if database connection fails
 # ======================================================================
 
 
@@ -35,10 +36,20 @@ out=" "
 # FUNCTIONS
 # ----------------------------------------------------------------------
 
+# exit with status unknown if a database query failed
+function _queryFailed(){
+    ph.setStatus "unknown"
+    echo "UNKNOWN: the database query failed"
+    echo "Quick troubleshooter:"
+    echo "* install a database user $myuser with command $(basename $0) -i"
+    echo "* check pg_hba.conf login fails if it set to ident (eg. change it to md5)"
+    ph.exit
+}
+
 # uninstall database user
 function _uninstall(){
-    echo UNINSTALL ...
-    su - postgres -c "psql -c \"DROP USER ${myuser};\""
+    echo "UNINSTALL ..."
+    su - postgres -c "psql -c \"DROP USER ${myuser};\"" 2>&1
     unset PGHOST
     unset PGUSER
     unset PGPASSWORD
@@ -47,7 +58,7 @@ function _uninstall(){
 
 # (re)install database user for monitoring
 function _install(){
-    echo INSTALLING ...
+    echo "INSTALLING ..."
     local pwlength=64
     echo "- check psql connection..."
     su postgres -c "psql -V postgres"
@@ -56,20 +67,18 @@ function _install(){
         exit 1
     fi
 
-    echo "- creating mysql user $myuser@localhost with random password ($pwlength chars)..."
+    echo "- creating database user $myuser with random password ($pwlength chars)..."
     mypw=$( head /dev/urandom | tr -dc A-Za-z0-9 | head -c $pwlength )
 
     SQL1="CREATE USER ${myuser} WITH PASSWORD '${mypw}' CONNECTION LIMIT 5";
     SQL2="GRANT pg_monitor TO ${myuser};"
     # SQL2="GRANT MONITOR_QUERIES TO ${myuser};"
-    su - postgres -c "psql -c \"${SQL1}\""
-    if [ $? -ne 0 ]; then
+    if ! su - postgres -c "psql -c \"${SQL1}\" 2>&1"; then
         echo "ERROR: psql command to create user failed."
         exit 1
     fi
     echo "- grant ..."
-    su - postgres -c "psql -c \"${SQL2}\""
-    if [ $? -ne 0 ]; then
+    if ! su - postgres -c "psql -c \"${SQL2}\" 2>&1" ; then
         echo "ERROR: psql command to grant permissions failed."
         # exit 1
     fi
@@ -86,8 +95,7 @@ export PGHOST=localhost
 # set default database because a user db won't be created
 export PGDATABASE=postgres
 EOF
-    ls -l $cfgfile
-    if [ $? -ne 0 ]; then
+    if ! ls -l $cfgfile; then
         echo "ERROR: creation of config file failed."
         exit 1
     fi
@@ -116,8 +124,8 @@ OPTIONS:
 PARAMETERS:
   -m  method; valid methods are:
       activity        running processes and queries
-      dbrows          Count of database row actions
       conflicts       Detected conflicts from pg_stat_database_conflicts
+      dbrows          Count of database row actions
       diskblock       Count of diskblocks physically read or coming from cache
       problems        Problems and troublemakers
       replication     Replication status (table output only)
@@ -135,7 +143,7 @@ EOH
 # param  string       database query
 function renderCounters(){
   local _query;      _query="$1"
-  local _out;        _out=$( psql -c "${_query}")
+  local _out;        _out=$( psql -c "${_query} 2>&1") || _queryFailed
   local _iCounter;   typeset -i _iCounter=0
   local _header;     _header=$( echo "${_out}" | head -1 | tr -d ' ')
   local _data;       _data=$( echo "${_out}" | head -3 | tail -1 | tr -d ' ')
@@ -200,8 +208,7 @@ fi
 
 
 # --- check installation
-grep $myuser $cfgfile >/dev/null 2>/dev/null
-if [ $? -ne 0 ]; then
+if ! grep $myuser $cfgfile >/dev/null 2>&1; then
     ph.abort "PSQL access not possible yet. You need to install the monitoring user first: as root execute `basename $0` -i"
 fi
 
@@ -213,7 +220,7 @@ sMode=$(ph.getValueWithParam '' m "$@")
 
 case "${sMode}" in
     "activity")
-        _out=$( psql -c "select pid,usename,state,query,backend_type,backend_start from pg_stat_activity" )
+        _out=$( psql -c "select pid,usename,state,query,backend_type,backend_start from pg_stat_activity" ) || _queryFailed
         typeset -i iQTotal;  iQTotal=$(  tail -1 <<< "$_out"  | cut -f 1 -d ' ' | tr -d '(' )
         typeset -i iQActive; iQActive=$( awk '{ print $5 }' <<< "$_out" | grep -c "active"  )
 
@@ -260,7 +267,7 @@ case "${sMode}" in
             from pg_stat_database "
           ;;
     "replication")
-        _out=$( psql -c "select * from pg_stat_replication" )
+        _out=$( psql -c "select * from pg_stat_replication" ) || _queryFailed
         if tail -1 <<< "$_out"  | grep "(0 rows)" >/dev/null ; then
             descr="No data in pg_stat_replication - this is no slave."
             out=""