clickhouse-config.xml

<?xml version="1.0"?>
<clickhouse>
    <!-- Network Configuration -->
    <http_port>8123</http_port>
    <tcp_port>9000</tcp_port>
    <interserver_http_port>9009</interserver_http_port>
    
    <!-- Allow access from any host -->
    <listen_host>0.0.0.0</listen_host>
    
    <!-- Memory and Performance Settings -->
    <max_memory_usage>10000000000</max_memory_usage>
    <max_memory_usage_for_user>10000000000</max_memory_usage_for_user>
    
    <!-- Buffer Settings for Batch Inserts -->
    <background_pool_size>16</background_pool_size>
    <background_schedule_pool_size>16</background_schedule_pool_size>
    
    <!-- MergeTree Settings -->
    <merge_tree>
        <!-- How many rows to buffer before writing a part -->
        <min_rows_for_compact_part>10000</min_rows_for_compact_part>
        <min_rows_for_wide_part>100000</min_rows_for_wide_part>
        
        <!-- Part size settings -->
        <min_bytes_for_wide_part>10000000</min_bytes_for_wide_part>
        <max_bytes_to_merge_at_min_space_in_pool>1048576</max_bytes_to_merge_at_min_space_in_pool>
        <max_bytes_to_merge_at_max_space_in_pool>161061273600</max_bytes_to_merge_at_max_space_in_pool>
        
        <!-- Merge settings -->
        <parts_to_delay_insert>300</parts_to_delay_insert>
        <parts_to_throw_insert>600</parts_to_throw_insert>
        <max_delay_to_insert>1</max_delay_to_insert>
        
        <!-- Storage settings -->
        <max_part_loading_threads>4</max_part_loading_threads>
        <max_parts_in_total>100000</max_parts_in_total>
    </merge_tree>
    
    <!-- Storage Configuration -->
    <storage_configuration>
        <disks>
            <default>
                <path>/var/lib/clickhouse/</path>
            </default>
        </disks>
        <policies>
            <default>
                <volumes>
                    <default>
                        <disk>default</disk>
                    </default>
                </volumes>
            </default>
        </policies>
    </storage_configuration>
    
    <!-- Query Settings -->
    <max_execution_time>300</max_execution_time>
    <max_threads>8</max_threads>
    
    <!-- Settings for CDC-style Updates -->
    <default_profile>
        <use_minimalistic_part_header_in_zookeeper>1</use_minimalistic_part_header_in_zookeeper>
        <load_balancing>random</load_balancing>
        <max_partitions_per_insert_block>100</max_partitions_per_insert_block>
        <max_insert_block_size>1048576</max_insert_block_size>
        <input_format_parallel_parsing>1</input_format_parallel_parsing>
    </default_profile>
    
    <!-- Compression Settings -->
    <compression>
        <case>
            <min_part_size>10000000</min_part_size>
            <min_part_size_ratio>0.01</min_part_size_ratio>
            <method>lz4</method>
        </case>
    </compression>

    <!-- Mark cache size for acceleration of repeated queries -->
    <mark_cache_size>5368709120</mark_cache_size>
    
    <!-- Buffer profile for batch inserts from Flink -->
    <profiles>
        <default>
            <max_insert_block_size>1048576</max_insert_block_size>
            <max_bytes_before_external_group_by>2147483648</max_bytes_before_external_group_by>
            <max_memory_usage>10000000000</max_memory_usage>
            <max_memory_usage_for_user>10000000000</max_memory_usage_for_user>
            <max_bytes_before_external_sort>2147483648</max_bytes_before_external_sort>
            <max_bytes_before_remerge_sort>2147483648</max_bytes_before_remerge_sort>
        </default>
    </profiles>

    <!-- Quotas -->
    <quotas>
        <default>
            <interval>
                <duration>3600</duration>
                <queries>0</queries>
                <errors>0</errors>
                <result_rows>0</result_rows>
                <read_rows>0</read_rows>
                <execution_time>0</execution_time>
            </interval>
        </default>
    </quotas>
</clickhouse>