0

I have 320 GB hard disk. I want to know how to partition it. Please give me specific details of space I need to use and their names. Can I use the name \home instead of any other name?

How much space I should use for the Ubuntu OS installation?

Kevin Bowen
  • 19,615
  • 55
  • 79
  • 83

1 Answers1

1

This is very use-case-specific. i.e. In order to be able to answer this, one needs to know how this system is going to be used.

I can give some examples from my own usage scenarios, though

Use case: Linux desktop, single boot

  • Swap => same size as system memory
  • / => 20GB
  • Unused partition => 20GB
  • /home => Remainder

Use case: Linux desktop, dual- or multi-boot

  • Swap => same size as system memory
  • / => 20GB
  • / (for second OS) => 20GB
  • /windows => 100GB
  • /home => Remainder

Use case: Linux web server

  • Swap => Same size as system memory. optional
  • / => Remainder

Use case: Linux database server

  • Swap => Same size as system memory. optional
  • / => 20GB
  • database data partition /data or /var/lib/mysql (for example) => remainder

For Linux desktops, I always reserve a second OS partition that I use to upgrade or test out new OSes. I can install the new OS on that partition and leave my main OS intact. If I like the new OS, I can switch to that as my main OS and mount my /home partition there. Boom, instant upgrade.

For multi-boot, the premise is similar. Note that Windows is much more disk-intensive than a typical Linux install. You really need to give modern Windows systems at least 100GB.

Server systems are highly variable. The partitioning really depends on how the server will be used, how much data it will store, how it is deployed, what backend infrastructure it runs on (e.g. AWS, VMWare, real machines per server, etc)

Finally, swap space on a modern Linux installation is optional. Some scenarios benefit greatly from it; some do not. For Linux desktops, I always include a swap partition. Not so much on my servers -- I just add more memory to those if I find that memory is becoming a resource constraint on them.

JDS
  • 111