oltp_point_selects
and oltp_index_updates
. In this document, all test cases involve 4 Select operations and 1 Update operation with the read/write ratio at 4:1.Type | Description |
Physical machine | High-IO edition where a single machine can support database instances with up to 488 GB memory and 6 TB disk |
Instance specification | Currently purchasable mainstream specification (please see the test cases below) |
Client configuration | 4-core CPU and 8 GB memory |
Number of clients | 1–6 (more clients need to be added as the configuration is upgraded) |
Network environment | Data center with 10-Gigabit connection and a network latency below 0.05 ms |
Environment load | Load on the machine where MySQL is installed is above 70% (for non-exclusive instances) |
CREATE TABLE `sbtest1` (`id` int(10) unsigned NOT NULL AUTO_INCREMENT,`k` int(10) unsigned NOT NULL DEFAULT '0',`c` char(120) NOT NULL DEFAULT '',`pad` char(60) NOT NULL DEFAULT '',PRIMARY KEY (`id`), KEY `k_1` (`k`)) ENGINE=InnoDB DEFAULT CHARSET=utf8;
id: 1k: 20106885c: 08566691963-88624912351-16662227201-46648573979-64646226163-77505759394-75470094713-41097360717-15161106334-50535565977pad: 63188288836-92351140030-06390587585-66802097351-4928296184
/root//sysbench/sysbench --mysql-host=xxxx --mysql-port=xxxx --mysql-user=xxx --mysql-password=xxx --mysql-db=test --mysql-table-engine=innodb --test=tests/db/oltp.lua --oltp_tables_count=20 --oltp-table-size=10000000 --rand-init=on prepare
--test=tests/db/oltp.lua
indicates to implement the OLTP test by calling the tests/db/oltp.lua script.--oltp_tables_count=20
indicates that the number of tables for testing is 20.--oltp-table-size=10000000
indicates that each testing table is populated with 10 million rows of data.--rand-init=on
indicates that each testing table is populated with random data./root//sysbench/sysbench --mysql-host=xxxx --mysql-port=xxx --mysql-user=xxx --mysql-password=xxx --mysql-db=test --test=/root/sysbench_for_z3/sysbench/tests/db/oltp.lua --oltp_tables_count=xx --oltp-table-size=xxxx --num-threads=xxx --oltp-read-only=off --rand-type=special --max-time=600 --max-requests=0 --percentile=99 --oltp-point-selects=4 run
--oltp_tables_count=20
indicates that the number of tables for testing is 20.--oltp-table-size=10000000
indicates that each testing table is populated with 10 million rows of data.--num-threads=128
indicates that the concurrent connections of clients for testing is 128.--oltp-read-only=off
indicates that the read-only testing model is disabled and the hybrid read/write model is used.--rand-type=special
indicates that the random model is specific.--max-time=1800
indicates the execution time of this test.--max-requests=0
indicates that no limit is imposed on the total number of requests and the test is executed according to max-time
.--percentile=99
indicates the sampling rate. Here, 99 means discarding 1% long requests of all the requests and taking the maximum value among the remaining 99% requests. The default value is 95%.--oltp-point-selects=4
indicates that the number of Select operations in the SQL testing command in the OLTP script is 4. The default value is 1.our_oltp.lua
which is modified to run four Select
operations and one Update
operation (index column) with the read/write ratio at 4:1.Memory | Storage Capacity | Number of Tables | Number of Rows | Data Set Size | Concurrence | Execution Time (Min) |
4 GB | 200 GB | 8 | 40 million | 76 GB | 128 | 30 |
8 GB | 200 GB | 15 | 40 million | 142 GB | 128 | 30 |
16 GB | 400 GB | 25 | 40 million | 238 GB | 128 | 30 |
32 GB | 700 GB | 25 | 40 million | 238 GB | 128 | 30 |
64 GB | 1 TB | 40 | 40 million | 378 GB | 256 | 30 |
96 GB | 1.5 TB | 40 | 40 million | 378 GB | 128 | 30 |
128 GB | 2 TB | 40 | 40 million | 378 GB | 128 | 30 |
244 GB | 3 TB | 60 | 40 million | 567 GB | 128 | 30 |
488 GB | 6 TB | 60 | 40 million | 567 GB | 128 | 30 |
488 GB (tuned) | 6 TB | 60 | 10 million | 140 GB | 128 | 30 |
Memory | Storage Capacity | Data Set | Number of Clients | Single-client Concurrence | QPS | TPS |
4 GB | 200 GB | 76 GB | 1 | 128 | 4,082 | 816 |
8 GB | 200 GB | 142 GB | 1 | 128 | 6,551 | 1,310 |
16 GB | 400 GB | 238 GB | 1 | 128 | 11,098 | 2,219 |
32 GB | 700 GB | 238 GB | 2 | 128 | 20,484 | 3,768 |
64 GB | 1 TB | 378 GB | 2 | 128 | 36,395 | 7,279 |
96 GB | 1.5 TB | 378 GB | 3 | 128 | 56,465 | 11,292 |
128 GB | 2 TB | 378 GB | 3 | 128 | 82,007 | 16,120 |
244 GB | 3 TB | 567 GB | 4 | 128 | 99,868 | 19,505 |
488 GB | 6 TB | 567 GB | 6 | 128 | 141,136 | 28,449 |
488 GB (tuned) | 6 TB | 140 GB | 6 | 128 | 245,408 | 47,201 |
Was this page helpful?