Arch Linux

Gullible Jones

Member

Registered: 2004-12-29

Posts: 4,863

Troubling ext2/3 limitations

Having just discovered the joys of directory-indexed ext3 filesystems, I've been thinking, "Gee, this would probably be a great FS for workstations."

But maybe not. According to Wikipedia, there is a 16 GB limit on filesize... Now, that may sound like a lot, but when you think about it, it ain't. In 5 years, we could be dealing with 10+ GB files.

It seems to me as though ext3 could do with a bit more attention...

jaboua

Member

Registered: 2005-11-05

Posts: 634

Re: Troubling ext2/3 limitations

Having just discovered the joys of directory-indexed ext3 filesystems, I've been thinking, "Gee, this would probably be a great FS for workstations."

But maybe not. According to Wikipedia, there is a 16 GB limit on filesize... Now, that may sound like a lot, but when you think about it, it ain't. In 5 years, we could be dealing with 10+ GB files.

It seems to me as though ext3 could do with a bit more attention...

In 5 years, ext3 will probably have evolved into tackling bigger files, maybe an ext4, who knows.

I don't think it's too much of an issue on my box, I have a 40G hdd ... However I have been using an 11GB file (harddisk image for qemu), but else than in that case, 700MB is usually the top.

Gullible Jones

Member

Registered: 2004-12-29

Posts: 4,863

Re: Troubling ext2/3 limitations

Ehh... I dunno. What worries me is that ext3 doesn't seem to be getting much attention compared to things like Reiser4, even though it's a very promising filesystem.

Dusty

Schwag Merchant

From: Medicine Hat, Alberta, Canada

Registered: 2004-01-18

Posts: 5,986

Website

Re: Troubling ext2/3 limitations

As usual, feel free to develop it further. ;-)

---

** Arch Linux Apparel and more ** Arch Linux stickers, laptop bags, jewelery ** I wrote a book on Python 3 ** and one on mental health **

jondkent

Member

From: London

Registered: 2005-09-13

Posts: 123

Re: Troubling ext2/3 limitations

Oh I dunno if it worth worrying about.  It's unlikely that you'll be using the same system in 5 years without at least rebuilding it at least once.  5 years is still the MTBF for hard disks last time I checked anyway.

Mind you ext3 does seem to have dropped off the radar, maybe becauase it not sexy and does what it says on the tin?

Jon

PJ

Member

From: Sweden

Registered: 2005-10-11

Posts: 602

Re: Troubling ext2/3 limitations

Having just discovered the joys of directory-indexed ext3 filesystems, I've been thinking, "Gee, this would probably be a great FS for workstations."

But maybe not. According to Wikipedia, there is a 16 GB limit on filesize... Now, that may sound like a lot, but when you think about it, it ain't. In 5 years, we could be dealing with 10+ GB files.

It seems to me as though ext3 could do with a bit more attention...

That was strange, when I was looking on Wikipedia I find a 2TiB limit for maximum filesize. 

Limits
Max file size:     2TiB
Max number of files:     Variable
Max filename size:     255 bytes
Max volume size:     32TiB

jaboua

Member

Registered: 2005-11-05

Posts: 634

Re: Troubling ext2/3 limitations

From here, it seems variable: http://en.wikipedia.org/wiki/Comparison_of_file_systems
maximum filesize:

16GiB to 2TiB4

If it's like I've heard about some other FSs, it's the blocksize which decides.

Gullible Jones

Member

Registered: 2004-12-29

Posts: 4,863

Re: Troubling ext2/3 limitations

Hmmm... Here's a kernel patch to make the filesize limit higher. I wonder... Are the limits of the filesystem determined by the kernel?

Mind you ext3 does seem to have dropped off the radar, maybe becauase it not sexy and does what it says on the tin?

For some reason, though, the tin usually doesn't mention directory indexing, which makes ext3 at least the equal of ReiserFS for most things.

codergeek42

Member

From: Anaheim, CA (USA)

Registered: 2005-06-03

Posts: 90

Website

Re: Troubling ext2/3 limitations

Ehh... I dunno. What worries me is that ext3 doesn't seem to be getting much attention compared to things like Reiser4, even though it's a very promising filesystem.

That's because Ext3 doesn't need much attention. It's solid, it's proven, it's highly scalable, it has a lot of nice features, and it works - extremely well. 

It's still being actively worked on though. In fact, cool things like delayed allocation and extents are being planned for Ext3 (possibly Ext4), and were discussed at the recent Ottowa Symposium

---

~Peter~

Gullible Jones

Member

Registered: 2004-12-29

Posts: 4,863

Re: Troubling ext2/3 limitations

Sounds cool... I'll stick with this FS, I think.

(I don't get the "highly scalable" bit though. That's its one weak spot really... Well, that and the relatively slow formatting of volumes.)

codergeek42

Member

From: Anaheim, CA (USA)

Registered: 2005-06-03

Posts: 90

Website

Re: Troubling ext2/3 limitations

"Highly scalable" in the sense that it can be used and tweaked to be an excellent FS for desktop/workstations with simple ~100 gig hard disks to  servers with many TB of high-end SCSI/SATA RAID for example or network storage.

---

~Peter~