This article is reprinted from the Jishu Community.
Jishu Column: Arm Server
Author:BoltLiu
This is a test report on enabling MySQL PGO. Correctly enabling MySQL PGO leads to significant performance improvements, with write performance increasing by about 13% and read performance increasing by about 20%.
Environment Setup
The testing environment is set up on an ARM server ECS requested from Aliyun, as shown in the table below:
MySQL PGO Compilation Steps
-
Add the parameter “-DFPROFILE_GENERATE=ON” during cmake, then compile:
$ cd mysql-server-8.0.33
$ mkdir build
$ cmake -DCMAKE_C_FLAGS="-g -O3 -march=native -mcpu=native -flto" -DCMAKE_CXX_FLAGS="-g -O3 -march=native -mcpu=native -flto" -DCMAKE_INSTALL_PREFIX=/mysql_data/mysql_8.0.33_gcc_11.3.0_profile -DWITH_BOOST=/build/boost_1_77_0 -DFPROFILE_GENERATE=ON ..
$ make -j $(nproc)
$ make install-
Start MySQL installed in /mysql_data/mysql_8.0.33_gcc_11.3.0_profile, run write and read performance tests from the client, and the profile data will be generated in the directory:
mysql-server-8.0.33/build-profile-data
-
Add the parameter “-DFPROFILE_USE=ON” and recompile MySQL:
$ cd mysql-server-8.0.33
$ rm rf build && mkdir build # make sure it's same build directory as in step 1, or it would fail to find the profile data under build-profile-data
$ cmake -DCMAKE_C_FLAGS="-g -O3 -march=native -mcpu=native -flto" -DCMAKE_CXX_FLAGS="-g -O3 -march=native -mcpu=native -flto" -DCMAKE_INSTALL_PREFIX=/mysql_data/mysql_8.0.33_gcc_11.3.0_pgo -DWITH_BOOST=/build/boost_1_77_0 -DFPROFILE_USE=ON ..
$ make -j $(nproc)
$ make installTest Results
-
Restart the server machine
-
Start the non-PGO version of MySQL
-
Run write tests
-
Run read tests
-
Repeat steps 1 to 4 a total of 3 times
-
Restart the server machine
-
Start the PGO version of MySQL
-
Run write tests
-
Run read tests
-
Repeat steps 6 to 9 a total of 3 times
Write Test Results
Throughput:
events/s (eps): 7837.0387
time elapsed: 300.0977s
total number of events: 2351876
Throughput:
events/s (eps): 7616.9932
time elapsed: 300.0844s
total number of events: 2285740
Throughput:
events/s (eps): 7893.9496
time elapsed: 300.0817s
total number of events: 2368829
Read Test Results
The following are the results of 3 rounds of read tests for the non-PGO version:
Throughput:
events/s (eps): 3768.8060
time elapsed: 300.1503s
total number of events: 1131208
Throughput:
events/s (eps): 3688.1464
time elapsed: 300.1504s
total number of events: 1106998
Throughput:
events/s (eps): 3774.9087
time elapsed: 300.1509s
total number of events: 1133042
PGO Version
Write Test Results (+13.4%, +16.7%, +11.8%)
The following are the write test results for the 3 rounds of the PGO version, compared to the non-PGO version, with performance improvements of 13.4%, 16.7%, and 11.8% respectively:
Throughput:
events/s (eps): 8891.5023
time elapsed: 300.0943s
total number of events: 2668288
Throughput:
events/s (eps): 8892.7030
time elapsed: 300.0876s
total number of events: 2668589
Throughput:
events/s (eps): 8831.1063
time elapsed: 300.0857s
total number of events: 2650088
Read Test Results (+25.9%, +20.9%, +16.4%)
The following are the read test results for the 3 rounds of the PGO version, compared to the non-PGO version, with performance improvements of 25.9%, 20.9%, and 16.4% respectively:
Throughput:
events/s (eps): 4746.7576
time elapsed: 300.1492s
total number of events: 1424735
Throughput:
events/s (eps): 4460.4811
time elapsed: 300.1489s
total number of events: 1338808
Throughput:
events/s (eps): 4395.7754
time elapsed: 300.1699s
total number of events: 1319479
Copyright belongs to the original author. If there is any infringement, please contact for deletion.
END
关于安芯教育
安芯教育是聚焦AIoT（人工智能+物联网）的创新教育平台，提供从中小学到高等院校的贯通式AIoT教育解决方案。
安芯教育依托Arm技术，开发了ASC（Arm智能互联）课程及人才培养体系。已广泛应用于高等院校产学研合作及中小学STEM教育，致力于为学校和企业培养适应时代需求的智能互联领域人才。