pgpool2.git / summary / log / commit / refs

commit    ea4df313e3cd60100113e2a9ee72440623748a72
Author:   Tatsuo Ishii <ishii@postgresql.org>
Date:     Mon Apr 09 05:45:37 2018 +0000

    Add new regression test for node 0 is down.
    
    test case 1: node 0 is already down before pgpool starts.
    test case 2: node 0 goes down after pgpool starts.
    test case 3: node 0 goes down and DISALLOW_TO_FAILOVER flag is set after pgpool starts.


 .../regression/tests/017.node_0_is_down/test.sh    | 141 +++++++++++++++++++++
 1 file changed, 141 insertions(+)

diff --git a/src/test/regression/tests/017.node_0_is_down/test.sh b/src/test/regression/tests/017.node_0_is_down/test.sh
new file mode 100755
index 000000000..dc7442190
--- /dev/null
+++ b/src/test/regression/tests/017.node_0_is_down/test.sh
@@ -0,0 +1,141 @@
+#!/usr/bin/env bash
+#-------------------------------------------------------------------
+# test script for the case when node 0 is down (and DISALLOW_FAILOVER is being set).
+#
+source $TESTLIBS
+TESTDIR=testdir
+PSQL=$PGBIN/psql
+PG_CTL=$PGBIN/pg_ctl
+export PGDATABASE=test
+
+for mode in s r n
+do
+	rm -fr $TESTDIR
+	mkdir $TESTDIR
+	cd $TESTDIR
+
+	echo "===== testing mode: $mode ======"
+
+	# create test environment
+	echo -n "creating test environment..."
+	$PGPOOL_SETUP -m $mode -n 2 || exit 1
+	echo "done."
+
+	source ./bashrc.ports
+
+	#
+	# Test case 1: when pgpool starts node 0 is already down and
+	# pool_status recorded the fact.
+	# Only streaming replication mode failed with read-only transaction.
+	# Other mode should succeed.
+	#
+	# edit startall so that it does not use "-D" with pgpool.
+	sed -i 's/pgpool -D/pgpool/' startall
+
+	# set down status for node 0.
+	echo -e "down \nup" > log/pgpool_status
+	
+	./startall
+
+	export PGPORT=$PGPOOL_PORT
+
+	wait_for_pgpool_startup
+
+	$PSQL -c "show pool_nodes" postgres
+	# try to issue write query
+	$PSQL -c "create table t1(i int)" postgres  > psql_status 2>&1
+	if [ $? != 0 ]
+	then
+	    if [ $mode = "s" ]
+	    then
+		grep -i read-only psql_status >/dev/null
+		if [ $? != 0 ]
+		then
+		    echo -n "Writing query in streamin replication mode failed due to other than read-only transaction in test case 1. reason: "
+		    cat psql_status
+		    ./shutdownall
+		    exit 1
+		fi
+	    else
+		echo -n "Writing query in mode $mode failed in test case 1. reason: "
+		cat psql_status
+		./shutdownall
+		exit 1
+	    fi
+	else
+	    $PSQL -c "drop table t1" postgres
+	fi
+	echo "Writing query in mode $mode succeeded in test case 1"
+
+	./shutdownall
+
+	#
+	# Test case 2: server starts then node 0 is shutting down.
+	# In all mode this test should succeed.
+	#
+	# remove pgpool_status to trigger failover.
+	rm log/pgpool_status
+	# set down request file
+	./startall
+	wait_for_pgpool_startup
+	# shutdown node 0
+	pg_ctl -D data0 -m f stop
+	wait_for_pgpool_startup
+	$PSQL -c "show pool_nodes" postgres
+	# try to issue write query
+	$PSQL -c "create table t1(i int)" postgres > psql_status 2>&1
+	if [ $? != 0 ]
+	then
+	    echo -n "Writing query in mode $mode failed in test case 2. reason: "
+	    cat psql_status
+	else
+	    echo "Writing query in mode $mode succeeded in test case 2"
+	fi
+
+	./shutdownall
+
+	$PSQL -c "drop table t1" postgres
+
+	#
+	# Test case 3: in addition 2 set DISALLOW_TO_FAILOVER flag to node 0.
+	#
+
+	# Set DISALLOW_TO_FAILOVER flag to node 0
+	echo "backend_flag0 = DISALLOW_TO_FAILOVER" >> etc/pgpool.conf
+	# remove pgpool_status to trigger failover.
+	rm log/pgpool_status
+	# set down request file
+	./startall
+	wait_for_pgpool_startup
+	# shutdown node 0
+	pg_ctl -D data0 -m f stop
+	wait_for_pgpool_startup
+	$PSQL -c "show pool_nodes" postgres
+	# try to issue write query expecting error: executing failover on backend.
+	$PSQL -c "create table t1(i int)" postgres > psql_status 2>&1
+	if [ $? != 0 ]
+	then
+	    grep -i "executing failover on backend" psql_status >/dev/null
+	    if [ $? != 0 ]
+	    then
+		echo -n "Writing query in mode $mode failed due to other than \"executing failover on backend\" in test case 3. reason: "
+		cat psql_status
+		./shutdownall
+		exit 1
+	    else
+		echo "Writing query in mode $mode failed by \"executing failover on backend\" as expected in test case 3"
+	    fi
+	else
+	    echo "Writing query succeeded in test case 3 mode $mode. Suspicious sccess."
+	    ./shutdownall
+	    exit 1
+	fi
+
+	$PSQL -c "drop table t1" postgres
+
+	./shutdownall
+
+	cd ..
+done
+
+exit 0


[parent: 0616c4f5de9e]