Skip to main content

Star-Schema-Benchmark

Star Schema Benchmark

Star Schema Benchmark(SSB) is a lightweight performance test set in the data warehouse scenario. SSB provides a simplified star schema data based on TPC-H, which is mainly used to test the performance of multi-table JOIN query under star schema. In addition, the industry usually flattens SSB into a wide table model (Referred as: SSB flat) to test the performance of the query engine, refer to Clickhouse.

This document mainly introduces the performance of Doris on the SSB 100G test set.

Note 1: The standard test set including SSB usually has a large gap with the actual business scenario, and some tests will perform parameter tuning for the test set. Therefore, the test results of the standard test set can only reflect the performance of the database in a specific scenario. It is recommended that users use actual business data for further testing.

Note 2: The operations involved in this document are all performed in the Ubuntu Server 20.04 environment, and CentOS 7 as well.

With 13 queries on the SSB standard test data set, we conducted a comparison test based on Apache Doris 1.2.0-rc01, Apache Doris 1.1.3 and Apache Doris 0.15.0 RC04 versions.

On the SSB flat wide table, the overall performance of Apache Doris 1.2.0-rc01 has been improved by nearly 4 times compared with Apache Doris 1.1.3, and nearly 10 times compared with Apache Doris 0.15.0 RC04.

On the SQL test with standard SSB, the overall performance of Apache Doris 1.2.0-rc01 has been improved by nearly 2 times compared with Apache Doris 1.1.3, and nearly 31 times compared with Apache Doris 0.15.0 RC04.

1. Hardware Environment

Number of machines4 Tencent Cloud Hosts (1 FE, 3 BEs)
CPUAMD EPYC™ Milan (2.55GHz/3.5GHz) 16 Cores
Memory64G
Network Bandwidth7Gbps
DiskHigh-performance Cloud Disk

2. Software Environment

  • Doris deployed 3BEs and 1FE;
  • Kernel version: Linux version 5.4.0-96-generic (buildd@lgw01-amd64-051)
  • OS version: Ubuntu Server 20.04 LTS 64-bit
  • Doris software versions: Apache Doris 1.2.0-rc01, Apache Doris 1.1.3 and Apache Doris 0.15.0 RC04
  • JDK: openjdk version "11.0.14" 2022-01-18

3. Test Data Volume

SSB Table NameRowsAnnotation
lineorder600,037,902Commodity Order Details
customer3,000,000Customer Information
part1,400,000Parts Information
supplier200,000Supplier Information
date2,556Date
lineorder_flat600,037,902Wide Table after Data Flattening

4. Test Results

We use Apache Doris 1.2.0-rc01, Apache Doris 1.1.3 and Apache Doris 0.15.0 RC04 for comparative testing. The test results are as follows:

QueryApache Doris 1.2.0-rc01(ms)Apache Doris 1.1.3 (ms)Doris 0.15.0 RC04 (ms)
Q1.12090250
Q1.2101030
Q1.33070120
Q2.190360900
Q2.2903401,020
Q2.360260770
Q3.11605501,710
Q3.280290670
Q3.390240550
Q3.4202030
Q4.11404801,250
Q4.250240400
Q4.330200330
Total8803,1508,030

ssb_v11_v015_compare

Interpretation of Results

  • The data set corresponding to the test results is scale 100, about 600 million.
  • The test environment is configured as the user's common configuration, with 4 cloud servers, 16-core 64G SSD, and 1 FE, 3 BEs deployment.
  • We select the user's common configuration test to reduce the cost of user selection and evaluation, but the entire test process will not consume so many hardware resources.

5. Standard SSB Test Results

Here we use Apache Doris 1.2.0-rc01, Apache Doris 1.1.3 and Apache Doris 0.15.0 RC04 for comparative testing. In the test, we use Query Time(ms) as the main performance indicator. The test results are as follows:

QueryApache Doris 1.2.0-rc01 (ms)Apache Doris 1.1.3 (ms)Doris 0.15.0 RC04 (ms)
Q1.14018350
Q1.23010080
Q1.3207080
Q2.135094020,680
Q2.232075018,250
Q2.330072014,760
Q3.16502,15022,190
Q3.22605108,360
Q3.32204506,200
Q3.46070160
Q4.18401,48024,320
Q4.24605606,310
Q4.361066010,170
Total4,1608,478131,910

ssb_12_11_015

Interpretation of Results

  • The data set corresponding to the test results is scale 100, about 600 million.
  • The test environment is configured as the user's common configuration, with 4 cloud servers, 16-core 64G SSD, and 1 FE 3 BEs deployment.
  • We select the user's common configuration test to reduce the cost of user selection and evaluation, but the entire test process will not consume so many hardware resources.

6. Environment Preparation

Please first refer to the [official documentation](. /install/install-deploy.md) to install and deploy Apache Doris first to obtain a Doris cluster which is working well(including at least 1 FE 1 BE, 1 FE 3 BEs is recommended).

The scripts mentioned in the following documents are stored in the Apache Doris codebase: ssb-tools

7. Data Preparation

7.1 Download and Install the SSB Data Generation Tool.

Execute the following script to download and compile the ssb-dbgen tool.

sh build-ssb-dbgen.sh

After successful installation, the dbgen binary will be generated under the ssb-dbgen/ directory.

7.2 Generate SSB Test Set

Execute the following script to generate the SSB dataset:

sh gen-ssb-data.sh -s 100 -c 100

Note 1: Check the script help via sh gen-ssb-data.sh -h.

Note 2: The data will be generated under the ssb-data/ directory with the suffix .tbl. The total file size is about 60GB and may need a few minutes to an hour to generate.

Note 3: -s 100 indicates that the test set size factor is 100, -c 100 indicates that 100 concurrent threads generate the data of the lineorder table. The -c parameter also determines the number of files in the final lineorder table. The larger the parameter, the larger the number of files and the smaller each file.

With the -s 100 parameter, the resulting dataset size is:

TableRowsSizeFile Number
lineorder600,037,90260GB100
customer3,000,000277M1
part1,400,000116M1
supplier200,00017M1
date2,556228K1

7.3 Create Table

7.3.1 Prepare the doris-cluster.conf File.

Before import the script, you need to write the FE’s ip port and other information in the doris-cluster.conf file.

The file location is at the same level as load-ssb-dimension-data.sh.

The content of the file includes FE's ip, HTTP port, user name, password and the DB name of the data to be imported:

export FE_HOST="xxx"
export FE_HTTP_PORT="8030"
export FE_QUERY_PORT="9030"
export USER="root"
export PASSWORD='xxx'
export DB="ssb"

7.3.2 Execute the Following Script to Generate and Create the SSB Table:

sh create-ssb-tables.sh

Or copy the table creation statements in create-ssb-tables.sql and create-ssb-flat-table.sql and then execute them in the MySQL client.

The following is the lineorder_flat table build statement. Create the lineorder_flat table in the above create-ssb-flat-table.sh script, and perform the default number of buckets (48 buckets). You can delete this table and adjust the number of buckets according to your cluster scale node configuration, so as to obtain a better test result.

CREATE TABLE `lineorder_flat` (
`LO_ORDERDATE` date NOT NULL COMMENT "",
`LO_ORDERKEY` int(11) NOT NULL COMMENT "",
`LO_LINENUMBER` tinyint(4) NOT NULL COMMENT "",
`LO_CUSTKEY` int(11) NOT NULL COMMENT "",
`LO_PARTKEY` int(11) NOT NULL COMMENT "",
`LO_SUPPKEY` int(11) NOT NULL COMMENT "",
`LO_ORDERPRIORITY` varchar(100) NOT NULL COMMENT "",
`LO_SHIPPRIORITY` tinyint(4) NOT NULL COMMENT "",
`LO_QUANTITY` tinyint(4) NOT NULL COMMENT "",
`LO_EXTENDEDPRICE` int(11) NOT NULL COMMENT "",
`LO_ORDTOTALPRICE` int(11) NOT NULL COMMENT "",
`LO_DISCOUNT` tinyint(4) NOT NULL COMMENT "",
`LO_REVENUE` int(11) NOT NULL COMMENT "",
`LO_SUPPLYCOST` int(11) NOT NULL COMMENT "",
`LO_TAX` tinyint(4) NOT NULL COMMENT "",
`LO_COMMITDATE` date NOT NULL COMMENT "",
`LO_SHIPMODE` varchar(100) NOT NULL COMMENT "",
`C_NAME` varchar(100) NOT NULL COMMENT "",
`C_ADDRESS` varchar(100) NOT NULL COMMENT "",
`C_CITY` varchar(100) NOT NULL COMMENT "",
`C_NATION` varchar(100) NOT NULL COMMENT "",
`C_REGION` varchar(100) NOT NULL COMMENT "",
`C_PHONE` varchar(100) NOT NULL COMMENT "",
`C_MKTSEGMENT` varchar(100) NOT NULL COMMENT "",
`S_NAME` varchar(100) NOT NULL COMMENT "",
`S_ADDRESS` varchar(100) NOT NULL COMMENT "",
`S_CITY` varchar(100) NOT NULL COMMENT "",
`S_NATION` varchar(100) NOT NULL COMMENT "",
`S_REGION` varchar(100) NOT NULL COMMENT "",
`S_PHONE` varchar(100) NOT NULL COMMENT "",
`P_NAME` varchar(100) NOT NULL COMMENT "",
`P_MFGR` varchar(100) NOT NULL COMMENT "",
`P_CATEGORY` varchar(100) NOT NULL COMMENT "",
`P_BRAND` varchar(100) NOT NULL COMMENT "",
`P_COLOR` varchar(100) NOT NULL COMMENT "",
`P_TYPE` varchar(100) NOT NULL COMMENT "",
`P_SIZE` tinyint(4) NOT NULL COMMENT "",
`P_CONTAINER` varchar(100) NOT NULL COMMENT ""
) ENGINE=OLAP
DUPLICATE KEY(`LO_ORDERDATE`, `LO_ORDERKEY`)
COMMENT "OLAP"
PARTITION BY RANGE(`LO_ORDERDATE`)
(PARTITION p1 VALUES [('0000-01-01'), ('1993-01-01')),
PARTITION p2 VALUES [('1993-01-01'), ('1994-01-01')),
PARTITION p3 VALUES [('1994-01-01'), ('1995-01-01')),
PARTITION p4 VALUES [('1995-01-01'), ('1996-01-01')),
PARTITION p5 VALUES [('1996-01-01'), ('1997-01-01')),
PARTITION p6 VALUES [('1997-01-01'), ('1998-01-01')),
PARTITION p7 VALUES [('1998-01-01'), ('1999-01-01')))
DISTRIBUTED BY HASH(`LO_ORDERKEY`) BUCKETS 48
PROPERTIES (
"replication_num" = "1",
"colocate_with" = "groupxx1",
"in_memory" = "false",
"storage_format" = "DEFAULT"
);

7.4 Import data

We use the following command to complete all data import of SSB test set and SSB FLAT wide table data synthesis and then import into the table.

 sh bin/load-ssb-data.sh -c 10

-c 5 means start 10 concurrent threads to import (5 by default). In the case of a single BE node, the lineorder data generated by sh gen-ssb-data.sh -s 100 -c 100 will also generate the data of the ssb-flat table in the end. If more threads are enabled, the import speed can be accelerated. But it will cost extra memory.

Notes.

  1. To get faster import speed, you can add flush_thread_num_per_store=5 in be.conf and then restart BE. This configuration indicates the number of disk writing threads for each data directory, 2 by default. Larger data can improve write data throughput, but may increase IO Util. (Reference value: 1 mechanical disk, with 2 by default, the IO Util during the import process is about 12%. When it is set to 5, the IO Util is about 26%. If it is an SSD disk, it is almost 0%) .

  2. The flat table data is imported by 'INSERT INTO ... SELECT ... '.

7.5 Checking Imported data

select count(*) from part;
select count(*) from customer;
select count(*) from supplier;
select count(*) from date;
select count(*) from lineorder;
select count(*) from lineorder_flat;

The amount of data should be consistent with the number of rows of generated data.

TableRowsOrigin SizeCompacted Size(1 Replica)
lineorder_flat600,037,90259.709 GB
lineorder600,037,90260 GB14.514 GB
customer3,000,000277 MB138.247 MB
part1,400,000116 MB12.759 MB
supplier200,00017 MB9.143 MB
date2,556228 KB34.276 KB

7.6 Query Test

7.6.1 SSB FLAT Test for SQL

--Q1.1
SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
FROM lineorder_flat
WHERE LO_ORDERDATE >= 19930101 AND LO_ORDERDATE <= 19931231 AND LO_DISCOUNT BETWEEN 1 AND 3 AND LO_QUANTITY < 25;
--Q1.2
SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
FROM lineorder_flat
WHERE LO_ORDERDATE >= 19940101 AND LO_ORDERDATE <= 19940131 AND LO_DISCOUNT BETWEEN 4 AND 6 AND LO_QUANTITY BETWEEN 26 AND 35;

--Q1.3
SELECT SUM(LO_EXTENDEDPRICE * LO_DISCOUNT) AS revenue
FROM lineorder_flat
WHERE weekofyear(LO_ORDERDATE) = 6 AND LO_ORDERDATE >= 19940101 AND LO_ORDERDATE <= 19941231 AND LO_DISCOUNT BETWEEN 5 AND 7 AND LO_QUANTITY BETWEEN 26 AND 35;

--Q2.1
SELECT SUM(LO_REVENUE), (LO_ORDERDATE DIV 10000) AS YEAR, P_BRAND
FROM lineorder_flat WHERE P_CATEGORY = 'MFGR#12' AND S_REGION = 'AMERICA'
GROUP BY YEAR, P_BRAND
ORDER BY YEAR, P_BRAND;

--Q2.2
SELECT SUM(LO_REVENUE), (LO_ORDERDATE DIV 10000) AS YEAR, P_BRAND
FROM lineorder_flat
WHERE P_BRAND >= 'MFGR#2221' AND P_BRAND <= 'MFGR#2228' AND S_REGION = 'ASIA'
GROUP BY YEAR, P_BRAND
ORDER BY YEAR, P_BRAND;

--Q2.3
SELECT SUM(LO_REVENUE), (LO_ORDERDATE DIV 10000) AS YEAR, P_BRAND
FROM lineorder_flat
WHERE P_BRAND = 'MFGR#2239' AND S_REGION = 'EUROPE'
GROUP BY YEAR, P_BRAND
ORDER BY YEAR, P_BRAND;

--Q3.1
SELECT C_NATION, S_NATION, (LO_ORDERDATE DIV 10000) AS YEAR, SUM(LO_REVENUE) AS revenue
FROM lineorder_flat
WHERE C_REGION = 'ASIA' AND S_REGION = 'ASIA' AND LO_ORDERDATE >= 19920101 AND LO_ORDERDATE <= 19971231
GROUP BY C_NATION, S_NATION, YEAR
ORDER BY YEAR ASC, revenue DESC;

--Q3.2
SELECT C_CITY, S_CITY, (LO_ORDERDATE DIV 10000) AS YEAR, SUM(LO_REVENUE) AS revenue
FROM lineorder_flat
WHERE C_NATION = 'UNITED STATES' AND S_NATION = 'UNITED STATES' AND LO_ORDERDATE >= 19920101 AND LO_ORDERDATE <= 19971231
GROUP BY C_CITY, S_CITY, YEAR
ORDER BY YEAR ASC, revenue DESC;

--Q3.3
SELECT C_CITY, S_CITY, (LO_ORDERDATE DIV 10000) AS YEAR, SUM(LO_REVENUE) AS revenue
FROM lineorder_flat
WHERE C_CITY IN ('UNITED KI1', 'UNITED KI5') AND S_CITY IN ('UNITED KI1', 'UNITED KI5') AND LO_ORDERDATE >= 19920101 AND LO_ORDERDATE <= 19971231
GROUP BY C_CITY, S_CITY, YEAR
ORDER BY YEAR ASC, revenue DESC;

--Q3.4
SELECT C_CITY, S_CITY, (LO_ORDERDATE DIV 10000) AS YEAR, SUM(LO_REVENUE) AS revenue
FROM lineorder_flat
WHERE C_CITY IN ('UNITED KI1', 'UNITED KI5') AND S_CITY IN ('UNITED KI1', 'UNITED KI5') AND LO_ORDERDATE >= 19971201 AND LO_ORDERDATE <= 19971231
GROUP BY C_CITY, S_CITY, YEAR
ORDER BY YEAR ASC, revenue DESC;

--Q4.1
SELECT (LO_ORDERDATE DIV 10000) AS YEAR, C_NATION, SUM(LO_REVENUE - LO_SUPPLYCOST) AS profit
FROM lineorder_flat
WHERE C_REGION = 'AMERICA' AND S_REGION = 'AMERICA' AND P_MFGR IN ('MFGR#1', 'MFGR#2')
GROUP BY YEAR, C_NATION
ORDER BY YEAR ASC, C_NATION ASC;

--Q4.2
SELECT (LO_ORDERDATE DIV 10000) AS YEAR,S_NATION, P_CATEGORY, SUM(LO_REVENUE - LO_SUPPLYCOST) AS profit
FROM lineorder_flat
WHERE C_REGION = 'AMERICA' AND S_REGION = 'AMERICA' AND LO_ORDERDATE >= 19970101 AND LO_ORDERDATE <= 19981231 AND P_MFGR IN ('MFGR#1', 'MFGR#2')
GROUP BY YEAR, S_NATION, P_CATEGORY
ORDER BY YEAR ASC, S_NATION ASC, P_CATEGORY ASC;

--Q4.3
SELECT (LO_ORDERDATE DIV 10000) AS YEAR, S_CITY, P_BRAND, SUM(LO_REVENUE - LO_SUPPLYCOST) AS profit
FROM lineorder_flat
WHERE S_NATION = 'UNITED STATES' AND LO_ORDERDATE >= 19970101 AND LO_ORDERDATE <= 19981231 AND P_CATEGORY = 'MFGR#14'
GROUP BY YEAR, S_CITY, P_BRAND
ORDER BY YEAR ASC, S_CITY ASC, P_BRAND ASC;

7.6.2 SSB Standard Test for SQL

--Q1.1
SELECT SUM(lo_extendedprice * lo_discount) AS REVENUE
FROM lineorder, dates
WHERE
lo_orderdate = d_datekey
AND d_year = 1993
AND lo_discount BETWEEN 1 AND 3
AND lo_quantity < 25;
--Q1.2
SELECT SUM(lo_extendedprice * lo_discount) AS REVENUE
FROM lineorder, dates
WHERE
lo_orderdate = d_datekey
AND d_yearmonth = 'Jan1994'
AND lo_discount BETWEEN 4 AND 6
AND lo_quantity BETWEEN 26 AND 35;

--Q1.3
SELECT
SUM(lo_extendedprice * lo_discount) AS REVENUE
FROM lineorder, dates
WHERE
lo_orderdate = d_datekey
AND d_weeknuminyear = 6
AND d_year = 1994
AND lo_discount BETWEEN 5 AND 7
AND lo_quantity BETWEEN 26 AND 35;

--Q2.1
SELECT SUM(lo_revenue), d_year, p_brand
FROM lineorder, dates, part, supplier
WHERE
lo_orderdate = d_datekey
AND lo_partkey = p_partkey
AND lo_suppkey = s_suppkey
AND p_category = 'MFGR#12'
AND s_region = 'AMERICA'
GROUP BY d_year, p_brand
ORDER BY p_brand;

--Q2.2
SELECT SUM(lo_revenue), d_year, p_brand
FROM lineorder, dates, part, supplier
WHERE
lo_orderdate = d_datekey
AND lo_partkey = p_partkey
AND lo_suppkey = s_suppkey
AND p_brand BETWEEN 'MFGR#2221' AND 'MFGR#2228'
AND s_region = 'ASIA'
GROUP BY d_year, p_brand
ORDER BY d_year, p_brand;

--Q2.3
SELECT SUM(lo_revenue), d_year, p_brand
FROM lineorder, dates, part, supplier
WHERE
lo_orderdate = d_datekey
AND lo_partkey = p_partkey
AND lo_suppkey = s_suppkey
AND p_brand = 'MFGR#2239'
AND s_region = 'EUROPE'
GROUP BY d_year, p_brand
ORDER BY d_year, p_brand;

--Q3.1
SELECT
c_nation,
s_nation,
d_year,
SUM(lo_revenue) AS REVENUE
FROM customer, lineorder, supplier, dates
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_orderdate = d_datekey
AND c_region = 'ASIA'
AND s_region = 'ASIA'
AND d_year >= 1992
AND d_year <= 1997
GROUP BY c_nation, s_nation, d_year
ORDER BY d_year ASC, REVENUE DESC;

--Q3.2
SELECT
c_city,
s_city,
d_year,
SUM(lo_revenue) AS REVENUE
FROM customer, lineorder, supplier, dates
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_orderdate = d_datekey
AND c_nation = 'UNITED STATES'
AND s_nation = 'UNITED STATES'
AND d_year >= 1992
AND d_year <= 1997
GROUP BY c_city, s_city, d_year
ORDER BY d_year ASC, REVENUE DESC;

--Q3.3
SELECT
c_city,
s_city,
d_year,
SUM(lo_revenue) AS REVENUE
FROM customer, lineorder, supplier, dates
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_orderdate = d_datekey
AND (
c_city = 'UNITED KI1'
OR c_city = 'UNITED KI5'
)
AND (
s_city = 'UNITED KI1'
OR s_city = 'UNITED KI5'
)
AND d_year >= 1992
AND d_year <= 1997
GROUP BY c_city, s_city, d_year
ORDER BY d_year ASC, REVENUE DESC;

--Q3.4
SELECT
c_city,
s_city,
d_year,
SUM(lo_revenue) AS REVENUE
FROM customer, lineorder, supplier, dates
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_orderdate = d_datekey
AND (
c_city = 'UNITED KI1'
OR c_city = 'UNITED KI5'
)
AND (
s_city = 'UNITED KI1'
OR s_city = 'UNITED KI5'
)
AND d_yearmonth = 'Dec1997'
GROUP BY c_city, s_city, d_year
ORDER BY d_year ASC, REVENUE DESC;

--Q4.1
SELECT /*+SET_VAR(parallel_fragment_exec_instance_num=4, enable_vectorized_engine=true, batch_size=4096, enable_cost_based_join_reorder=true, enable_projection=true) */
d_year,
c_nation,
SUM(lo_revenue - lo_supplycost) AS PROFIT
FROM dates, customer, supplier, part, lineorder
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_partkey = p_partkey
AND lo_orderdate = d_datekey
AND c_region = 'AMERICA'
AND s_region = 'AMERICA'
AND (
p_mfgr = 'MFGR#1'
OR p_mfgr = 'MFGR#2'
)
GROUP BY d_year, c_nation
ORDER BY d_year, c_nation;

--Q4.2
SELECT /*+SET_VAR(parallel_fragment_exec_instance_num=2, enable_vectorized_engine=true, batch_size=4096, enable_cost_based_join_reorder=true, enable_projection=true) */
d_year,
s_nation,
p_category,
SUM(lo_revenue - lo_supplycost) AS PROFIT
FROM dates, customer, supplier, part, lineorder
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_partkey = p_partkey
AND lo_orderdate = d_datekey
AND c_region = 'AMERICA'
AND s_region = 'AMERICA'
AND (
d_year = 1997
OR d_year = 1998
)
AND (
p_mfgr = 'MFGR#1'
OR p_mfgr = 'MFGR#2'
)
GROUP BY d_year, s_nation, p_category
ORDER BY d_year, s_nation, p_category;

--Q4.3
SELECT /*+SET_VAR(parallel_fragment_exec_instance_num=2, enable_vectorized_engine=true, batch_size=4096, enable_cost_based_join_reorder=true, enable_projection=true) */
d_year,
s_city,
p_brand,
SUM(lo_revenue - lo_supplycost) AS PROFIT
FROM dates, customer, supplier, part, lineorder
WHERE
lo_custkey = c_custkey
AND lo_suppkey = s_suppkey
AND lo_partkey = p_partkey
AND lo_orderdate = d_datekey
AND s_nation = 'UNITED STATES'
AND (
d_year = 1997
OR d_year = 1998
)
AND p_category = 'MFGR#14'
GROUP BY d_year, s_city, p_brand
ORDER BY d_year, s_city, p_brand;