Member Avatar for Joncamp

I would like to know if anyone has bench marked MySQL server, compared to Oracle. I have developed a database that will sort 1,000,000 records in less then 22 seconds (100 byte record sorting on 2 fields, and 1.3Ghz cpu).

  • 2 Contributors
  • 2 Replies
  • 115 Views
  • 1 Week Discussion Span
  • Latest Post Latest Post by d66

Dani AI

Generated

Short summary: has asserted a faster database implementation than MySQL and asked for a URL/details, but the thread does not include a reproducible test plan or configuration. Below is a concise, practical checklist for anyone who wants to make a verifiable MySQL vs Oracle comparison and for what to post so others can reproduce and validate results.

Performance differences almost always come from workload and environment, not just "the engine." Important variables to state and control include DB and OS versions, storage (SSD vs spindle), filesystem and mount options, schema and indexes, concurrency pattern (single client vs many), transaction isolation/durability settings, storage engine (InnoDB/MyISAM/etc.), and full server configuration files. Without matching those, comparisons are not apples-to-apples.

Recommended benchmark workflow:

  • Choose a workload (OLTP vs OLAP) and a standard harness such as TPC suites or open tools like HammerDB (https://www.hammerdb.com/) or sysbench (https://github.com/akopytov/sysbench).
  • Document environment: CPU, RAM, disk model, OS/kernel, DB version, and attach full config files.
  • Generate a realistic dataset and perform warm-up runs so caches are comparable.
  • Run multiple iterations, capture throughput and latency percentiles (50/95/99), and collect OS counters (iostat/vmstat), DB metrics, and query plans (EXPLAIN / AWR / TKPROF).
  • Share schema, SQL, and the exact test script so results are reproducible.

Quick tuning and gotchas: size buffer pools to keep working set in memory, be explicit about commit/durability tradeoffs, watch for sorts spilling to disk, and remember Oracle and MySQL implement concurrency and parallelism differently (so feature parity matters when comparing). Also isolate the test host from noisy processes and run several trials.

What to post here to get useful feedback: exact DB versions, hardware specs, full configuration files, the workload definition or tool script, representative SQL + EXPLAIN output, raw run logs (throughput/latency) and system metrics. For benchmark standards and methodology see the TPC site (https://www.tpc.org/) and the tools above.

Member Avatar for Joncamp

I have a database which I developed that is 10 times faster then MySQL! No joke!

Member Avatar for d66

I have a database which I developed that is 10 times faster then MySQL! No joke!

Hey John, I've been seeing your post all over the web about database that you have developed. Are you going to put up a web address? Are you looking for employees? Seriously!

Dustin

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.