#!/usr/bin/env bash # Startup test: Start clicon daemon in the (four) different startup modes # (init, none, running, or startup) # The dbs and files are setup as follows: # - The example reset_state callback adds "lo" interface # - An extra xml configuration file starts with an "extra" interface # - running db starts with a "run" interface # - startup db starts with a "start" interface # There is also an "invalid" XML and a "broken" XML # There are two steps, first run through everything OK # Then try with invalid and broken XML and ensure the backend quits and all is untouched # Magic line must be first in script (see README.md) s="$_" ; . ./lib.sh || if [ "$s" = $0 ]; then exit 0; else return 0; fi # Which format to use as datastore format internally : ${format:=xml} APPNAME=example cfg=$dir/conf_startup.xml cat < $cfg $cfg ietf-netconf:startup /usr/local/share/clixon $IETFRFC clixon-example $APPNAME /usr/local/lib/$APPNAME/backend /usr/local/lib/$APPNAME/netconf /usr/local/lib/$APPNAME/restconf /usr/local/lib/$APPNAME/cli /usr/local/lib/$APPNAME/clispec /usr/local/var/$APPNAME/$APPNAME.sock /usr/local/var/$APPNAME/$APPNAME.pidfile $dir 0 init false $format EOF # Create running-db containin the interface "run" OK runvar='runex:ethtrue' # Create startup-db containing the interface "startup" OK startvar='startupex:ethtrue' # extra OK extravar='extraex:ethtrue' # invalid (contains ), but OK XML syntax invalidvar='invalidex:ethtrue' # Broken XML (contains ) brokenvar='brokenex:ethtrue' # Create a pre-set running, startup and (extra) config. # The configs are identified by an interface called run, startup, extra. # Depending on startup mode (init, none, running, or startup) # expect different output of an initial get-config of running testrun(){ mode=$1 rdb=$2 # running db at start sdb=$3 # startup db at start edb=$4 # extradb at start exprun=$5 # expected running_db after startup sudo rm -f $dir/*_db echo "$rdb" > $dir/running_db echo "$sdb" > $dir/startup_db echo "$edb" > $dir/extra_db if [ $BE -ne 0 ]; then # Bring your own backend # kill old backend (if any) new "kill old backend" sudo clixon_backend -zf $cfg if [ $? -ne 0 ]; then err fi new "start backend -f $cfg -s $mode -c $dir/extra_db" start_backend -s $mode -f $cfg -c $dir/extra_db new "waiting" wait_backend else new "Restart backend as eg follows: -Ff $cfg -s $mode -c $dir/extra_db # $BETIMEOUT s" sleep $BETIMEOUT fi new "Startup test for $mode mode, check running" expecteof "$clixon_netconf -qf $cfg" 0 ']]>]]>' "^$exprun]]>]]>$" new "Startup test for $mode mode, check candidate" expecteof "$clixon_netconf -qf $cfg" 0 ']]>]]>' "^$exprun]]>]]>$" new "Startup test for $mode mode, check startup is untouched" expecteof "$clixon_netconf -qf $cfg" 0 ']]>]]>' "^$sdb]]>]]>$" new "Kill backend" # Check if premature kill pid=$(pgrep -u root -f clixon_backend) if [ -z "$pid" ]; then err "backend already dead" fi # kill backend stop_backend -f $cfg } # testrun # The backend should fail with 255 and all db:s should be unaffected testfail(){ mode=$1 rdb=$2 # running db at start sdb=$3 # startup db at start edb=$4 # extradb at start sudo rm -f $dir/*_db echo "$rdb" > $dir/running_db echo "$sdb" > $dir/startup_db echo "$edb" > $dir/extra_db # kill old backend (if any) new "kill old backend" sudo clixon_backend -zf $cfg if [ $? -ne 0 ]; then err fi new "start backend -f $cfg -s $mode -c $dir/extra_db" ret=$(start_backend -1 -s $mode -f $cfg -c $dir/extra_db 2> /dev/null) r=$? if [ $r -ne 255 ]; then err "Unexpected retval" $r fi # permission kludges sudo chmod 666 $dir/running_db sudo chmod 666 $dir/startup_db new "Checking running unchanged" ret=$(diff $dir/running_db <(echo "$rdb")) if [ $? -ne 0 ]; then err "$rdb" "$ret" fi new "Checking startup unchanged" ret=$(diff $dir/startup_db <(echo "$sdb")) if [ $? -ne 0 ]; then err "$sdb" "$ret" fi new "Checking extra unchanged" ret=$(diff $dir/extra_db <(echo "$edb")) if [ $? -ne 0 ]; then err "$edb" "$ret" fi } # 1. Try different modes on OK running/startup/extra # Init mode: delete running and reload from scratch (just extra) testrun init "$runvar" "$startvar" "$extravar" "$extravar" # None mode: do nothing, running remains testrun none "$runvar" "$startvar" "$extravar" "$runvar" # Running mode: keep running but load also extra testrun running "$runvar" "$startvar" "$extravar" 'extraex:ethtruerunex:ethtrue' # Startup mode: scratch running, load startup with extra on top testrun startup "$runvar" "$startvar" "$extravar" 'extraex:ethtruestartupex:ethtrue' # 2. Try different modes on Invalid running/startup/extra WITHOUT failsafe # ensure all db:s are unchanged after failure. # Valgrind backend tests make no sense in backend crash tests if [ $valgrindtest -ne 2 ]; then new "Test invalid running in running mode" testfail running "$invalidvar" "$startvar" "$extravar" new "Run invalid startup in startup mode" testfail startup "$runvar" "$invalidvar" "$extravar" new "Test broken running in running mode" testfail running "$brokenvar" "$startvar" "$extravar" new "Run broken startup in startup mode" testfail startup "$runvar" "$brokenvar" "$extravar" fi rm -rf $dir