MariaDB vs. MySQL: A Performance Comparison 112
Nerval's Lobster writes "MariaDB is a fork of the MySQL source code, split off in the wake of concerns over what Oracle would do with MySQL licensing. In addition to its role as a 'drop-in replacement' for MySQL, MariaDB also includes some new features that (some claim) make it better than MySQL. Jeff Cogswell compares MySQL and MariaDB and suggests (in his opinion) that there's 'more than enough reason to ditch MySQL and switch over to MariaDB and stay there.' Why? While he breaks down MariaDB's new features and thinks many of them aren't that fantastic, and while MariaDB's performance isn't that much better than that of MySQL ('MariaDB's performance appears a bit better on multi-core machines, but I strongly suspect that one could tweak MySQL to match'), the questions over Oracle and MySQL licensing give him pause. 'MariaDB shows every indication that it will be around for quite awhile, while you can't really say the same of Oracle's MySQL,' he writes. 'Free-and-open MySQL competes with Oracle's proprietary and extremely competitive tools. That alone is grounds for concern — will Oracle do something to impede MySQL's development?'"
"will Oracle do something to impede MySQL's dev?" (Score:5, Interesting)
Re:Never thought I'd see FUD from Open Source (Score:5, Interesting)
My favorite part is that the article is titled "MariaDB vs. MySQL: A Performance Comparison", but since the performance is almost identical they spend most of the summary talking about ideological differences. I guess "MariaDB vs. MySQL: An Ideological Comparison" didn't have the same ring to it.
Re:When the incompatabilities ? (Score:4, Interesting)
But there is no good reason to switch besides hypothetical future situations. If one is already on either product, just stay there until something happens... so long as they are both still compatible.
The performance issues noted in the article doesn't make any sense to me. From the article:
One common result of not coding these correctly is you’ll start out seeing an improvement in the first 8 or 16 threads, and after that you won’t get nearly the hoped-for improvement. When you see that problem, it means there’s likely trouble with the algorithms. (And this will be the case with either hyperthreads or hardware threads.) That’s what we’re seeing here with the MySQL benchmarks. To me, that’s an indication of trouble with MySQL scaling, and should be a cause for concern. MariaDB also has a slight problem in the same benchmark as the performance goes down slightly, but only barely; I would surmise that this isn’t a problem with the parallel algorithms.
...but the graph at the top of the article does NOT show that! What benchmark is he referring to? The one at the top of the article shows MySQL 5.5.29 performing almost exactly the same as both versions of MariaDB that were tested. MySQL 5.6.10 performed a little different, which is because... well, we don't know due to lack of information there (he even says it may not have been compiled correctly, and a few other possible reasons).
If MariaDB does perform noticably better on many core machines (> 32), I'd be interested to know that, and that could be a justifiable reason to consider it.Even so, why go through that effort? And if you're running >32 cores on your DB, chances are it's a large DB and would take a fair bit of time to smoothly cut over... not really worth it IMO.
Re:Never thought I'd see FUD from Open Source (Score:5, Interesting)
MariaDB is taking the MySQL code via the GPL and then building on top of it with new code. Those changes are all having their copyright assigned to MariaDB, and in some cases the GPL will also require a public release. Eventually MariaDB is expected to have a non-trivial set of improvements, and the copyright ownership of all the new code will be to MariaDB. That allows selling the combination of GPL core plus some explicitly owned private code, the exact same way MySQL was sold to Sun.
This is the same scam that let Monty cash out once already, using the work of open source contributors who assigned their copyright to his original company. No reason he can't do it again, if people are gullible enough to fall for it twice.
MariaDB is more quickly that MySQL: it's true!!! (Score:5, Interesting)
- http://www.saltos.net/portal/en/wiki/75/why-use-mariadb-instead-of-mysql.htm [saltos.net]
Josep Sanz.
The SaltOS project.
Comment removed (Score:5, Interesting)
make a good switch (Score:2, Interesting)
There's been reason to switch from MySQL for 13 years: Postgres.