datastore/tests/performance/join_lazy_indexing.xml

<test>
    <substitutions>
        <substitution>
            <name>lazy_indexing</name>
            <values>
                <value>0</value>
                <value>3</value>
            </values>
        </substitution>
        <substitution>
            <name>large_table</name>
            <values>
                <value>large_int</value>
                <value>large_str</value>
            </values>
        </substitution>
    </substitutions>

    <create_query>CREATE TABLE large_int (a Int32, b Int32, c Int32, d Int32, e Int32, f Int32, g Int32, h Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE large_str (a Int32, b Int32, c Int32, s1 String, s2 String, s3 String, s4 String, s5 String) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small1 (a Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small2 (b Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small3 (c Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small1_dup (a Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small2_dup (b Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>
    <create_query>CREATE TABLE small3_dup (c Int32) ENGINE = MergeTree ORDER BY tuple()</create_query>

    <fill_query>INSERT INTO large_int SELECT number % 10000, number % 12500, number % 11000, number * 2, number * 3, number * 4, number * 5, number * 6 FROM numbers(100000000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO large_str SELECT number % 10000, number % 12500, number % 11000, toString(number), toString(number * 2), toString(number * 3), toString(number * 4), toString(number * 5) FROM numbers(100000000) ORDER BY rand()</fill_query>

    <fill_query>INSERT INTO small1 SELECT number FROM numbers(8000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small2 SELECT number FROM numbers(10000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small3 SELECT number FROM numbers(9000) ORDER BY rand()</fill_query>

    <fill_query>INSERT INTO small1_dup SELECT number FROM numbers(8000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small1_dup SELECT number FROM numbers(8000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small2_dup SELECT number FROM numbers(10000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small2_dup SELECT number FROM numbers(10000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small3_dup SELECT number FROM numbers(9000) ORDER BY rand()</fill_query>
    <fill_query>INSERT INTO small3_dup SELECT number FROM numbers(9000) ORDER BY rand()</fill_query>

    <!-- 1 join, no duplicates -->
    <query>SELECT * FROM {large_table} l JOIN small1 s ON l.a = s.a LIMIT 80000000, 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>
    <query>SELECT * FROM {large_table} l JOIN small1 s ON l.a = s.a ORDER BY l.a LIMIT 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>
    <query>SELECT * FROM {large_table} l JOIN small1 s ON l.a = s.a ORDER BY l.a LIMIT 1000 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>

    <!-- 1 join, with duplicates -->
    <query>SELECT * FROM {large_table} l JOIN small1_dup s ON l.a = s.a LIMIT 80000000, 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>
    <query>SELECT * FROM {large_table} l JOIN small1_dup s ON l.a = s.a ORDER BY l.a LIMIT 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>
    <query>SELECT * FROM {large_table} l JOIN small1_dup s ON l.a = s.a ORDER BY l.a LIMIT 1000 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash'</query>

    <!-- 3 joins, no duplicates -->
    <query>SELECT * FROM {large_table} l JOIN small1 s1 ON l.a = s1.a JOIN small2 s2 ON l.b = s2.b JOIN small3 s3 ON l.c = s3.c LIMIT 80000000, 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1 s1 ON l.a = s1.a JOIN small2 s2 ON l.b = s2.b JOIN small3 s3 ON l.c = s3.c ORDER BY l.a LIMIT 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1 s1 ON l.a = s1.a JOIN small2 s2 ON l.b = s2.b JOIN small3 s3 ON l.c = s3.c ORDER BY l.a LIMIT 1000 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1 s1 ON l.a = s1.a JOIN small2 s2 ON l.b = s2.b JOIN small3 s3 ON l.c = s3.c FORMAT NULL SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>

    <!-- 3 joins, with duplicates -->
    <query>SELECT * FROM {large_table} l JOIN small1_dup s1 ON l.a = s1.a JOIN small2_dup s2 ON l.b = s2.b JOIN small3_dup s3 ON l.c = s3.c LIMIT 80000000, 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1_dup s1 ON l.a = s1.a JOIN small2_dup s2 ON l.b = s2.b JOIN small3_dup s3 ON l.c = s3.c ORDER BY l.a LIMIT 10 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1_dup s1 ON l.a = s1.a JOIN small2_dup s2 ON l.b = s2.b JOIN small3_dup s3 ON l.c = s3.c ORDER BY l.a LIMIT 1000 SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>
    <query>SELECT * FROM {large_table} l JOIN small1_dup s1 ON l.a = s1.a JOIN small2_dup s2 ON l.b = s2.b JOIN small3_dup s3 ON l.c = s3.c FORMAT NULL SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>

    <!-- 3 joins, mixed duplicates -->
    <query>SELECT * FROM {large_table} l JOIN small1 s1 ON l.a = s1.a JOIN small2_dup s2 ON l.b = s2.b JOIN small3_dup s3 ON l.c = s3.c FORMAT NULL SETTINGS query_plan_min_columns_for_join_lazy_indexing = {lazy_indexing}, join_algorithm = 'hash', query_plan_join_swap_table = false</query>

    <drop_query>DROP TABLE IF EXISTS large_int</drop_query>
    <drop_query>DROP TABLE IF EXISTS large_str</drop_query>
    <drop_query>DROP TABLE IF EXISTS small1</drop_query>
    <drop_query>DROP TABLE IF EXISTS small2</drop_query>
    <drop_query>DROP TABLE IF EXISTS small3</drop_query>
    <drop_query>DROP TABLE IF EXISTS small1_dup</drop_query>
    <drop_query>DROP TABLE IF EXISTS small2_dup</drop_query>
    <drop_query>DROP TABLE IF EXISTS small3_dup</drop_query>
</test>