EmergencyEMERGENCY? Get 24/7 Help Now!

Blog Poll: What Operating System Do You Run Your Production Database On?

 | June 8, 2017 |  Posted In: MongoDB, MySQL


blog pollIn this post, we’ll use a blog poll to find out what operating system you use to run your production database servers.

As databases grow to meet more challenges and expanding application demands, they must try and get the maximum amount of performance out of available resources. How they work with an operating system can affect many variables, and help or hinder performance. The operating system you use for your database can impact consumable choices (such as hardware and memory). The operating system you use can also impact your choice of database engine as well (or vice versa).

Please let us know what operating system you use to run your database. For this poll, we’re asking which operating system you use to actually run your production database server (not the base operating system).

If you’re running virtualized Linux on Windows, please select Linux as the OS used for development. Pick up to three that apply. Add any thoughts or other options in the comments section:

What operating system do you use to run your production database?

  • CentOS 7 (20%, 446 Votes)
  • Ubuntu 16.04 LTS (16%, 362 Votes)
  • CentOS 6 (12%, 268 Votes)
  • RHEL 7 (9%, 209 Votes)
  • Debian 8 (9%, 197 Votes)
  • Ubuntu 14.04 LTS (8%, 184 Votes)
  • Windows Server (5%, 105 Votes)
  • RHEL 6 (5%, 101 Votes)
  • Other (4%, 81 Votes)
  • AWS (RDS and Aurora) (4%, 81 Votes)
  • FreeBSD (3%, 67 Votes)
  • Debian 7 (2%, 55 Votes)
  • Ubuntu 12.04 LTS (or earlier) (1%, 29 Votes)
  • CentOS 5 (or earlier) (1%, 21 Votes)
  • Solaris (1%, 17 Votes)
  • RHEL 5 (or earlier) (0%, 9 Votes)
  • Debian 6 (or earlier) (0%, 8 Votes)

Total Voters: 1,722

Loading ... Loading ...

Thanks in advance for your responses – they will help the open source community determine how database environments are being deployed.



          • Not necessarily. You can use binary packages compiled on some other box (see https://wiki.gentoo.org/wiki/Binary_package_guide). Alternatively, you can compile locally but use distcc to make compilation less stressful for other running processes (see https://wiki.gentoo.org/wiki/Distcc).
            But in our case we found that compiling packages regular way (without distcc) with just MAKEOPTS=”-j1″ and PORTAGE_NICENESS=”19″ is quite acceptable. We do, however, use binary packages locally to quickly reinstall packages (into chroot, for example, or when you need to roll back a package to a previous version or previous combination of USE flags, etc.). Also, if you’re using CFQ disk scheduler you can consider using ionice (PORTAGE_IONICE_COMMAND).
            All in all, compiling packages is usually not as scary as it sounds.

          • Well back in my Gentoo days I usually had a bit of breakage when doing upgrades that’s why I wondered. I also had a few clients running Gentoo who basically gave up on updates at some point. It’s always a somewhat delicate process, nowadays I would generally separate the OS from the data and just cycle the OS image completely. I’m usually constrained by running in a virtualized environment however.

            Thanks for taking the time to answer.

Leave a Reply