• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

partitioning

B

bolovan

n00b
Joined
May 22, 2013
Messages
3
Hello,
I have one dilema
Currently I have 4x2Tb hard drives in raid6 in my server (software raid with mdadm).
In the future, when I will be out of space I will add more hard drives and grow the raid.
over the raid I have a lvm volume group with one big lv (logical volume). My is allow me to add up to 12x3.5" hard drives.
what I have on my hard drive:
/car - where I keep movies made with my camera from car. files between 700Mb and 4Gb. Now I have around 800Gb
/surveillance - my surveillance cameras. Around 800Gb
/Documentaries - around 200Gb
/sports: what I captured from my tv, like fromula 1, 24 hours le mans. Files are between 900Mb and 5Gb. Currently around 300Gb
/other video: around 200Gb
Each year I add 2Tb of data on raid
My dilema:
1)to keep everything on one big partition, and when I grow raid I will grow also the partition
2)keep separate partion (for example: /car: 1Tb, /surveillance 1Tb, /Documentaries: 300Gb, /sports: 500Gb, /other: 300Gb). And when I be out of space, just add some Gb to partition that is out of space (lvm knows how to resize the volume and file sistem in the same time).
Any pro and cons?
 
At work (where I have around 50TB of mdadm raid6 scattered on a 1/2 dozen servers) I used lvm ontop of mdadm raid6 for almost a decade to do similar to what you are trying. For us it was splitting up research projects (with 100s of GB of data to a few TB) into their own filesystems. Now I just use btrfs ontop of mdadm raid6 and use the subvolume feature. This is much simpler than using lvm. Although btrfs still comes with the expermental tag. I have tested quite a bit before moving to it on to it in spring of last year. I think that was at the 3.5 kernel although at this point most of my servers are running 3.9.X kernels with a few running 3.8.X.
 
I am using lvm+ext4 becouse are quite mature.
My setup is using debian with kernel 3.9.1
I am still puzzled how to use.
One of pro partitioning is the problem that I have now: I have default inode numbers...witch are around 233 milions for 4Tb volume...but I have only 23.000 used (the smallest file is around 4Mb). So, I have 68Gb waisted space. To change this...I have to format again the partition...but I don't have where to store 2.4Tb of data.
On the other side...resizing volumes too ofen (if I use small partitions) is lead to fragmentation of partition, and in case of crash, is hard to recover the partition
 
I am using lvm+ext4 becouse are quite mature.
Click to expand...

That was what I did for years before the move to btrfs.

On the other side...resizing volumes too ofen (if I use small partitions) is lead to fragmentation of partition,
Click to expand...

You can defrag a lvm if you have space. Its quite safe to do (it creates temporary lvm mirrors of the fragments as it moves them) however it takes forever.

http://bisqwit.iki.fi/source/lvm2defrag.html
 
Last edited:
Yes...but my dilema is still here...to have one big partition...or to make 4-5 small partitions?
In the next 4-5 years I am expecting to have around 8-10 hard drives with 12-16Tb of space
 
You must log in or register to reply here.
Back
Top