Why we make partitions in Linux?
 

Hi
I want to clear a point regarding to creating partitions when we are installing linux.normally users make / i.e root partition and /home
so some make /boot,/var,/usr,/swap and
/home
Kindly guide me technically why we make different kinds of partitions? If we doesn't make any partition what will happen? And what is the advantages of making partions.
Thanks
garden

Maintenance!!!

Once a bug filled my /var with logs and since i didn't have separated it from /root entire system collapsed. Ever since, I have /tmp and /var separated. Use to have /usr too, but not any more cause kde4 is in /opt now. /home is must.

For easy backup, if home is on a separate partition, you can reuse it when you reinstall.

Kind regards

Additionally, AFAIR /boot has to be on a physical partition. If you have LVM, than I think you need to have a separate /boot. Furthermore,usually your filesystem is ext3/4, whereas swap is something completely different, so it has to be on a separate partition.

All you really need is two partitions. i.e / and /swap (assuming normal desktop usage etc). There are many reasons to have more then two and the following links explain the advantages and disadvantages.

http://www.cyberciti.biz/tips/the-im...artitions.html
http://en.wikipedia.org/wiki/Disk_partitioning

Well good guidance.what is the purpose of /usr and /temp
Does all softwares install in /usr partion?

What don't you understand?

Kind regards

The Filesystem Hierarchy Standard (FHS) answers both questions and more (the answer to your second question is "no").

Way back in the dimdarks, when a 100 MB drive was a rare and expensive beast, computers would have several small drives and needed to be able to spread the system out between them. That's why the plethora of root directories as system administrators could easily use them to mount on whatever storage they had available. The drives also tended to be slow compared to modern drives so spreading things out was a means to distribute input and output to optimize system performance.

Nowadays, with RAID, cheap gigabyte drives, SSD's, and with software ideas such as LVM and the updates for BIOS replacements and the way disk space is defined, the needs and reasons have changed. (MBR vs GPT etc)

Even a separate swap partition no longer has benefits it used to and a swap file can work as well in most circumstances.

So we have updates to boot loaders to be able to find and boot the systems in ever more complex environments. Change is rapid and that makes it a pain to keep up with it but the benefits are in such things as not having to figure out what to put where like you did way back when.

The maintenance issue described above is probably best handled by a re-install, IMHO. If you have a 10 or 20 GB root partition and a separate home, it is very easy to refresh an install with a modern Linux system, update, reconfigure, and be back where you were in only a few minutes. at least for most of us with a bit of due care.

I am looking at awe at being able to install GRUB2 on a USB memory stick and being able select a CD Image to boot that has a system with an encapsulated compressed file system - and then have that system provide a reasonable responsiveness to get things done. There is a lot going on under the sheets and it just works - but then a 16 GB memory device, USB ports, and all the other hardware is pretty incredible too (at least compared to twenty or thirty years ago).