Tagged articles
sysbench
37 articles · Page 1 of 1
Adding three secondary indexes to a MySQL 8.0 orders table caused write P99 latency to jump from 15ms to 2.3s and TPS to drop from 4800 to 1100 during peak hours; the article details a reproducible forensic process using A/B testing, invisible indexes, and sysbench to prove the culprit index had zero read benefit and safely remove it.
A/B testingInnoDBMySQL 8.0
0 likes · 23 min read

This article presents a comprehensive guide on applying Linux Cgroups to isolate MySQL resources, covering CPU, memory, I/O, and network bandwidth controls, detailed configuration steps, performance testing with sysbench, and practical observations to ensure stable and predictable database behavior.
LinuxMySQLPerformance Tuning
0 likes · 47 min read

This article explains how to troubleshoot and resolve the MySQL "Too many connections" error by using GDB to modify the max_connections parameter on a running MySQL 5.7 instance without restarting, including step‑by‑step commands, sysbench load testing, and two practical methods.
MySQLdatabase tuninggdb
0 likes · 9 min read

Percona’s analysis of MySQL version adoption reveals a slow migration to 8.x, prompting extensive performance tests that compare default and tuned configurations of MySQL 5.7, 8.0.36, 8.4, Percona Server, and MariaDB using sysbench and TPC‑C benchmarks, with detailed results showing consistent performance regressions in newer releases.
MySQLPerconaTPC-C
0 likes · 11 min read


ITPUBMay 2, 2024 · Databases
This article explains how Profile‑Guided Optimization (PGO) can increase MySQL throughput by 30‑50% without changing application code, detailing the required compilation steps, necessary CMake flags, and real‑world benchmark results that demonstrate the performance gains and their limitations.
C++MySQLPGO
0 likes · 10 min read

An extensive benchmark compares MySQL 8.2 Community and Enterprise editions using sysbench and TPC‑C on identical EC2 instances, revealing that, under default configurations, the Enterprise edition offers no consistent performance advantage over the Community edition, with only occasional gains from optional plugins.
Community vs EnterpriseDatabase ConfigurationMySQL
0 likes · 10 min read

This article explains how to use MySQL's built‑in mysqlslap tool and the third‑party sysbench utility to benchmark MySQL performance, covering configuration changes, command‑line options, data preparation, execution of read/write and I/O tests, and interpretation of the resulting metrics.
Database BenchmarkLinuxMySQL
0 likes · 16 min read

This article explains how MySQL's offline_mode variable and the pt‑kill tool can be used by DBAs to place a server in maintenance mode, instantly dropping non‑privileged client connections, preventing new ones, and providing a safe way to perform maintenance or backups.
Database MaintenanceMySQLOFFLINE_MODE
0 likes · 8 min read

This article walks through generating a heavy‑load MySQL workload, using top, iostat, and iotop to pinpoint high disk I/O, mapping OS thread IDs back to MySQL sessions, and safely killing the offending query to relieve pressure.
LinuxMySQLPerformance Tuning
0 likes · 20 min read

This article explains how excessive Linux CPU context switches affect system performance and shows step‑by‑step how to monitor and analyze them using vmstat, pidstat, and sysbench, including interpreting voluntary versus involuntary switches and interrupt statistics.
Linuxcontext switchperformance
0 likes · 10 min read

This article examines why MySQL benchmark results often show low transactions per second on seemingly powerful servers, exploring hidden factors such as network bottlenecks, SSL overhead, and inappropriate sort_buffer_size settings, and provides practical troubleshooting steps and configuration tweaks to achieve optimal performance.
MySQLNetwork BottleneckSSL
0 likes · 10 min read

The article analyzes the instability of shared‑type virtual machines, introduces enterprise‑level instances with fixed CPU scheduling and NUMA topology, details the applied technologies such as CPU pinning, PCI‑Passthrough and multi‑queue NICs, and presents extensive sysbench and STREAM benchmark results that demonstrate superior isolation, stability and performance of enterprise instances over shared ones.
CPU pinningNUMAcloud computing
0 likes · 12 min read

This article demonstrates how to simulate accidental deletion of a MySQL InnoDB .ibd file, observe the impact on queries, recover the file using the file descriptor in /proc, restart MySQL, and verify that various DDL operations can detect the missing tablespace.
InnoDBMySQLdatabase troubleshooting
0 likes · 12 min read

The article explains that mysqlslap appears to produce smoother latency results than sysbench because mysqlslap reports metrics per test round rather than per individual SQL statement, leading to mis‑interpretation of Max/Avg/Min values, while sysbench can reveal per‑statement latency variations.
MySQLSQL latencybenchmarking
0 likes · 4 min read

This article investigates why MySQL encounters out‑of‑memory errors during sysbench stress testing with high concurrency, examines system memory metrics, explains Linux vm.overcommit_memory settings, and provides a practical solution to prevent OOM by adjusting kernel parameters.
LinuxMySQLOOM
0 likes · 11 min read

This article presents a systematic performance test of the DTLE data replication tool, using sysbench, tc, and iperf3 to simulate different network latency and bandwidth scenarios, and compares DTLE's replication delay and bandwidth usage against native MySQL replication across three test cases.
bandwidthdatabase-replicationiperf3
0 likes · 7 min read

This guide explains how to use MySQL’s built‑in mysqlslap tool and the third‑party sysbench utility to perform comprehensive database stress testing, covering configuration changes, command examples, result interpretation, and additional CPU and I/O benchmarks for accurate performance evaluation.
LinuxMySQLdatabase benchmarking
0 likes · 15 min read

This article presents a systematic benchmark of MySQL 5.7.22 and MySQL 8.0.15 using sysbench to measure TPS and QPS across read‑write, read‑only, and write‑only modes under two persistence configurations, analyzes the results, and provides a Bash script for reproducing the tests.
BenchmarkMySQLdatabase
0 likes · 9 min read

This guide explains how to configure MySQL for high‑concurrency testing, use the built‑in mysqlslap tool and the third‑party sysbench utility to benchmark read/write performance, CPU, I/O, and memory, and interpret the resulting metrics for database optimization.
Database BenchmarkLinuxMySQL
0 likes · 15 min read

This article explains how to perform MySQL performance benchmarking with the built‑in mysqlslap tool and the third‑party sysbench utility, covering configuration changes, command‑line options, data preparation, execution steps, and interpretation of the resulting metrics.
BenchmarkMySQLmysqlslap
0 likes · 15 min read

This tutorial explains MySQL benchmarking fundamentals and walks you through installing sysbench, configuring test parameters, preparing data, running the benchmark, analyzing results, and applying best‑practice recommendations to accurately assess database performance.
Database PerformanceLinuxMySQL
0 likes · 11 min read

The article evaluates how turning on binlog transaction compression in MySQL 8.0 (both 8.0.19 and 8.0.20) affects throughput and resource usage under a sysbench workload of 120 tables with 240 concurrent threads, showing roughly a 1% performance drop and a similar increase in CPU consumption.
CompressionMySQLbinlog
0 likes · 5 min read

Using sysbench, this study benchmarks MySQL 5.7.22 and MySQL 8.0.15 under read‑write, read‑only, and write‑only workloads with various sync_binlog and innodb_flush_log_at_trx_commit settings, revealing performance trends across different concurrency levels and highlighting that MySQL 5.7 often outperforms 8.0 in these tests.
MySQLQPSTPS
0 likes · 11 min read

This article details a performance comparison between UCloud's KuaiJie‑based UDB service and self‑built MySQL instances, covering environment setup, sysbench benchmarking, result analysis, and cost evaluation to guide optimal database instance selection.
Cloud DatabaseMySQLUDB
0 likes · 8 min read

Using sysbench oltp_read_write benchmarks on a 100 GB database with a 25 GB buffer pool, this article evaluates the impact of varying innodb_buffer_pool_instances (1, 2, 4, 8, 16, 32, 64) on throughput and variability, concluding that 64 instances yields the highest performance.
MySQLinnodb_buffer_pool_instancessysbench
0 likes · 7 min read

This article explains MySQL benchmark testing, the role and metrics of benchmarking, introduces sysbench, provides step‑by‑step installation and usage instructions, and offers practical tips for interpreting results and optimizing database performance in production environments.
MySQLbenchmarkingdatabase
0 likes · 11 min read

This guide explains how to use MySQL's built‑in mysqlslap tool and the third‑party sysbench utility to simulate concurrent client loads, adjust server settings, run read/write and I/O tests, and interpret the resulting performance metrics for database optimization.
IO testingMySQLSQL
0 likes · 14 min read

This article details a comprehensive experiment that evaluates MySQL's ability to sustain up to 100,000 simultaneous connections using Percona Server's thread‑pool plugin, sysbench clients, adjusted OS/network limits, and multiple IP addresses, presenting hardware specs, configuration steps, observed errors, performance metrics, and practical recommendations.
MySQLdatabase scalinghigh concurrency
0 likes · 8 min read

This article demonstrates a step‑by‑step experiment using sysbench and performance_schema to measure how different values of innodb_buffer_pool_instances affect buffer‑pool lock latency, analyzes the collected data, and discusses the observed trends and limitations.
MySQLinnodb_buffer_pool_instancesperformance_schema
0 likes · 5 min read

This article presents a detailed MySQL performance benchmark of Alibaba Cloud RDS, Tencent Cloud RDS, and AWS RDS using sysbench, analyzes configuration differences, shows results across mixed, read‑only, and no‑index update workloads, and offers practical recommendations for cloud database tuning.
Cloud RDSMySQLdatabase optimization
0 likes · 12 min read

A sysbench benchmark comparing MySQL 5.7.25 and MySQL 8.0.15 shows that latin1 charset favors 5.7, while utf8mb4 dramatically improves 8.0 performance, and that collations such as utf8mb4_unicode_ci can further affect query speed.
CharsetDatabase BenchmarkMySQL
0 likes · 4 min read

An in‑depth performance benchmark compares MySQL on CentOS 6 and CentOS 7 using sysbench with 10 tables of 5 million rows, revealing up to 40% higher QPS on CentOS 7, detailing test setup, results, and key configuration tweaks that dramatically improve throughput.
CentOSMySQLSQL
0 likes · 7 min read

The article analyzes why Sysbench‑driven MySQL connections time out when concurrency exceeds 5 000, tracing the issue to TCP SYN‑cookie processing and a full accept queue, and proposes configuration‑level mitigations.
Linux kernelMySQLSYN cookie
0 likes · 8 min read

This article translates and expands on Mark Callaghan’s sysbench benchmark, revealing that MySQL 5.6‑5.7 suffers the steepest QPS drop in low‑concurrency tests, compares configurations across versions, presents detailed charts, and concludes that most performance regressions are isolated to the 5.7 release.
BenchmarkMySQLQPS
0 likes · 8 min read

This article explains how Meituan‑Dianping leverages the open‑source tool Sysbench for repeatable database performance testing, unifies test parameters, optimizes MySQL settings such as slave_parallel_workers, extends functionality via Lua scripts, and reveals the tool’s high‑performance data‑collection techniques.
Database TestingLua scriptingMySQL
0 likes · 13 min read

This article presents a detailed benchmark comparing PostgreSQL 9.6 and MySQL 5.7 on identical high‑end hardware using pgbench and SysBench, describing the test methodology, configuration tweaks, performance results for point selects and OLTP workloads, and concluding that both systems scale to modern demanding workloads.
BenchmarkMySQLPostgreSQL
0 likes · 11 min read


ITPUBMay 10, 2016 · Databases
A performance test using sysbench on a dual‑node DL380 G7 MySQL 5.6 cluster shows that semi‑synchronous replication delivers roughly one‑third the TPS of asynchronous replication and highlights recovery and replication‑lag limitations that can be mitigated with sharding and parallel workers.
Database ReplicationMySQLSemi-sync Replication
0 likes · 6 min read
