What should be the disk space of the Ubuntu to be on safer side?

What should be the disk space of the Ubuntu to be on safer side??

Not sure I understand the question correctly.
I tend to use small system partitions, 12 GB on my server without GUI, 48 GB on laptops and desktops with GUI, and many additional installed packages.
Safe is, if the disks aren’t full, but don’t waste too much space.
I have my system partitions usually 50…70% full.
Check /var/log few times a week, as it may fill up the disk, if logrotate is not properly configured.
If you see a more or less constant space occupation there, you are safe.
Oh, forgot to mention, I symlink stuff in /var/lib, so mysql data or example doesn’t take space on the system partition.

I would suggest that you partition the drive to have 40G for (root) / partition and at least 100 GBs for a (home) /home partition if you are using the default setup just one partition then 140gbs. That will give you room for added programs and storage. But that all depends on what you want ubuntu to do for you.

1 Like

That’s what I deploy for Ubuntu Server… then another mountpoint for whatever data the requested users want… (e.g. at /opt, or /var/database, whatever)…

For desktops? I wouldn’t have a clue - because I don’t partition my desktop machines other than default (i.e. recommendation for /boot and then EVERYthing else including /home is on the same filesystem - I really detest micromanaging filesystems - and I don’t dual boot - so usually only have “/boot” and “/”)…

In most cases I usually have enough RAM I don’t need swap - but when I do, I usually go for a swapfile instead of swap partition (yeah I guess if I wanted to tweak every ounce of grunt - I’d probably be better off with a swap partition next to /boot [or even before boot - if it was a mechanical/magnetic hard disk]).

Side rant / thread hijack :
Note : got a bunch of systems (debian, RHEL, Ubuntu) I support where some penny-pinching m0r0n has decided to use some arbitrary scaling scheme that originated in the 1980’s where storage was INCREDIBLY expensive (i.e. more expensive than system administration labour) :

256 MB for /boot
8 GB for /
4 GB for /var (WTF??? this a-hole is continually filling up)

And I spend an inordinate amount time micromanaging storage in /var/ to keep these pezzo di merda running - and it’s worst when it happens to a “P2” (high priority) device at 3 am and I get a call!

Or I merrily plod along trying to patch them - until it fills up the measly 256 GB for /boot…

Who are these people? They need to be fired and blacklisted from EVER working in IT ever again!

These days storage is DIRT CHEAP (Moore’s law applies almost equally to storage as CPU grunt) and labour is the main cost and way more expensive than even SSD storage! Hence the drive towards automation and DevOps - because capitalism demands efficiency savings EVERY financial year!

e.g. I get woken up at 3 am on Sunday morning - and I always bill a minimum of 1 hour at double-time… I’ve no idea what we bill the customer, but it’s more than my employer pays me - imagine that happens 5 x per week 52 weeks a year… I don’t know the exact amount - but I’d guess it’s in the $1000’s of dollars… 12 months of SSD storage from a cloud provider would be considerably less… And some M0R0N built this crap thinking they were saving money???

1 Like