You are not logged in.
Hi
Since I'll be upgrading my home and office workstations soon, I've decided to tinker a little with my current "go-to" partition scheme to better serve my current project needs.
Let's start with the office workstation. Here's the current scheme:
[256GB SSD]
sda1 # Windows boot
sda2 128GB # Windows system
sda3 ext2 500MB /boot
sda4 # encrypted
vault
vg # LVM
lv_root ext4 28GB /
lv_swap swap 8GB swap
lv_opt ext4 6-8GB /opt
lv_var etx4 24GB /var
lv_tmp ext2 2GB /tmp
lv_home ext4 [free-10GB] /home
free
And here is my current concept:
[256GB SSD]
sda1 # Windows boot
sda2 128GB # Windows system
sda3 ext2 250-500MB /boot
sda4 # encrypted
vault
vg # LVM
lv_root ext4 26-30GB /
lv_swap swap 4-8GB swap
lv_var etx4 12-16GB /var
lv_varlib etx4? 24-36GB /var/lib
lv_vartmp etx4/reiserfs? 2-4GB /var/tmp
lv_tmp ext2? 2GB /tmp
lv_home ext4 [free-10GB] /home
free
Basically the whole concept behind the change is to maximize /var flexibility/performance with applications like VMs (KVM) and containers (Docker). Was thinking about substituting LVM model with btrfs subvolume model, but I'm not convinced with it's maturity for production (will be a good choice for home workstation do).
Questions
1. Is separating /var/lib a good idea in those terms?
2. If so, what fs would fit in?
3. Should I consider creating separate partitions for other locations?
4. Any other tweaks/ideas that might come in hand?
5. (bonus) Which filesystems would You recommend for better compilation performance (ReiserFS/Reiser4/tmpfs)?
Offline